Software Patents

Read Baase Chapter 4 sections 5-7 which includes patents

Some patent papers

1. Simpson Garfinkel, Patently Absurd, 1993
Garfinkel's article is pretty easy reading, pointing out some problems with software patents specifically. But maybe things were different in 1993?

2.Richard Stallman on Patents, 2002
Richard Stallman

Stallman is against software patents, of course. However, his case here is better than many open-source-related arguments; in fact, it is squarely aligned with the interests of software-development businesses.

3.Paul Graham, a computer scientist and one of the partners of the venture-capital firm Y Combinator, wrote a 2006 essay Are Software Patents Evil?

Paul Graham

Graham makes the following claim early on:

One thing I do feel pretty certain of is that if you're against software patents, you're against patents in general. Gradually our machines consist more and more of software. Things that used to be done with levers and cams and gears are now done with loops and trees and closures. There's nothing special about physical embodiments of control systems that should make them patentable, and the software equivalent not.

Is this true?

Graham also says,

Frankly, it surprises me how small a role patents play in the software business. It's kind of ironic, considering all the dire things experts say about software patents stifling innovation, but when one looks closely at the software business, the most striking thing is how little patents seem to matter.

But that paragraph is about software companies being sued by other software companies, and not "patent trolls".

Graham also makes some other claims, in particular some about the role of the patent system in business competition generally. Check out what he says about Reveal.


Baase 4e §4.5 / 5e §4.6

Pharmaceutical Patents
Sofware-patent issues
Obvious in Context
The Benson decision
The Flook decision
The Diehr decision
Some software-patent examples
Heckel paper
Patent Trolls
Bessen & Meurer
Business-method patents
Stallman essay
Graham essay
KSR v Teleflex
The Bilski Case
Mayo Labs v Prometheus
Abstract patents and CLS v Alice 
Lawsuit Venue 

Do patents help advance progress? or hinder it?

Patents are pretty clearly a market regulation with the sole goal of improving innovation in technology. Inventors don't "deserve" to profit from their ideas; we simply want to make sure they are motivated to continue. In other words, this is purely utilitarian.

Patents are intended to cover INVENTIONS rather than IDEAS. If you have an idea to sell hamburgers with salsa, or newspapers & beer together, or to create a website where people can post their own stuff, that's an IDEA. It can't be protected: everyone else is entitled to copy it freely.

If you decide that a certain vertical market can use an XML-based word processor, is that an invention or is that a business idea?

What do these have to do with computers? There are several issues about whether the patent system for software in fact helps anyone, and whether software patents fulfill their constitutional mandate to benefit society as a whole. These relate to the situation where patents are dealt with only by serious developers.

Currently in the US, patents last 20 years, and generally cannot be renewed or extended (but see below regarding pharmaceutical patents). During that period, the patent owner can enforce their patent, but patent litigation is relatively expensive. Formerly US patents lasted 17 years from the date of granting, but to comply with the WTO treaty on Trade-Related Aspects of Intellectual Property Rights (TRIPS) the term was changed to 20 years from application.

There is also an analogue to the copyright situation, where individuals can download software that may violate patents in some countries. This creates a situation somewhat related to file-sharing: end-users make the decision. Unlike file-sharing, there is no group analogous to the RIAA that is going after infringers.

Look at ubuntu software installation? What about MP3 players?

35 U.S.C. §101 (patent-eligibility law):

Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.

The meaning of "process" is critical here: does it mean any procedure or method? Or does it mean "industrial process"? Historically, it pretty clearly was intended to mean the latter.

Pharmaceutical patents

Pharmaceutical patents are sort of the poster child for Why Patents Are Good For Us (though Boldrin and Levine have argued against even pharmaceutical patents in their book). Here the patent system IS generally seen as effective at encouraging investment:

You might not at first really believe the second point above. But take a look at the Open Insulin Foundation, They are literally trying to create an open insulin-production process that can be implemented by individuals (though more likely by small groups).

Note that, despite these two points, pharmaceutical patents don't exactly encourage research on potential new drugs. New-drug research mostly comes from NIH-funded university research. Pharma patents primarily encourage investment in the FDA approval process (which is really the expensive part).

One weirdness involving pharmaceutical patents is that a patent can be "for the use of"; someone can, if drug X is in the public domain, patent the use of X to treat disease Y (this must be in some legal sense a "new" use of X). In practice this is not much of a problem, because generic manufacturers can still make and market X for its old purpose, and doctors can prescribe it for its new purpose. Such prescriptions are sometimes said to be "off-label"; they are an important way for drugs to get to people who will probably be helped by them, but for which no company has yet done clinical trials, and never will.

Some specific drugs:

Cancer monoclonal-antibody drugs

Cancers are caused by runaway cell replication, due to mutation. Traditional approaches to fighting cancer involved finding ways to kill cells that were dividing rapidly. Such treatment also affects cells that normally divide rapidly: hair roots and bone marrow, to name two groups.

However, mutations are ubiquitous; only some mutations lead to runaway replication. When this does happen, there is usually a specific genetic reason: the mutation unleashed some existing cellular mechanism to multiply. This has led to the idea that if we can identify a multiplicative mechanism, and find ways to shut it down, then we can attack cancers that utilize that mechanism. There is at least some evidence that many cancers share a relatively small common set of replication mechanisms.

These protein compounds that affect cellular replication are antibodies that are highly specific to a certain substrate; they have very narrow targets. The From

Another branch in targeted therapy is the increasing use of monoclonal antibodies in cancer therapy. Although monoclonal antibodies (immune proteins which can be selected to precisely bind to almost any target) have been around for decades, they were derived from mice and did not function particularly well when administered to humans, causing allergic reactions and being rapidly removed from circulation. "Humanization" of these antibodies (genetically transforming them to be as similar to a human antibody as possible) has allowed the creation of a new family of highly effective humanized monoclonal antibodies. Rituximab, a drug used to treat lymphomas, is a prime example. -- Wikipedia

The point is that some cancers can be specifically targeted by certain antibodies, because they have specific antibody receptors not present or not relevant in non-cancerous cells. The receptors involved tend to be very idiosyncratic. Antibodies are produced by specific immune-system cells; an antibody is monoclonal if it is (or was at one time) produced by one specific cell that was repeatedly cloned. Generally speaking, "monoclonal" means that the antibody has only a single receptor; that is, its activity is very focused.

None of these drugs would exist in the US marketplace if it were not for pharmaceutical patents.

On the other hand, the US Food and Drug Administration, which regulates new drugs, is arguably a massive government intrusion into the free market. Why shouldn't patent law intrude as well?

/gleevec: leukemias, stomach cancers. It is used to treat cancers where the cells involved have a specific receptor. Time magazine hailed it as the "magic bullet against cancer" in 2001, when it was approved, though that was an overbroad assessment.

It is the first member of a new class of agents that act by inhibiting particular tyrosine kinase enzymes, instead of non-specifically inhibiting rapidly dividing cells. -- Wikipedia

The idea is that only cancer cells make these particular enzymes (in particular the enzyme BCR-Abl), in a way closely related to the way the cells become cancerous.

Imatinib was the subject of a patent lawsuit in India, below.

rituximab/rituxan: binds to the white-blood-cell surface protein CD20. This makes those cells more visible to existing immune-system defenses. Used to treat leukemias/lymphomas, also some autoimmune diseases such as lupus

ibritumomab tiuxetan/Zevalin also binds to CD20; it is often tagged with a radioactive compound to deliver radiation therapy (that is, death) directly to the cancerous white blood cells.

cetuximab/Erbitux: metastatic colorectal cancer, head&neck cancers. Binds to EGFR receptor; it is an EGFR-inhibitor.

trastuzumab/Herceptin: breast cancer. Monoclonal antibody that interferes with HER2/neu receptor. In some breast cancers, the HER2 receptor is, as wikipedia put it, "stuck in the 'on' position".

crizotinib/Xalcori: attacks lung cancers triggered by a mutation in the ALK gene. The ALK gene encodes a protein that triggers the unlimited-cell-division process of the cancer.

vemurafenib/Zelboraf: blocks a protein that causes skin cells to replicate leading to melanoma.

ipilimumab/Yervoy: T-lymphocytes (a kind of white blood cell) attack certain cancer cells, but there are biological mechanisms in place that limit this action. Ipilimumab disables these limiting mechanisms, allowing the T-lymphocytes to work more effectively.

There are other drugs, too, of course. For some the patent-needed argument is just as strong; for others the situation is less clear.


There are very few new antibiotics, actually. The FDA used to require not just proof of effectiveness, but proof that the infection that was cured was in fact resistant to existing antibiotics. A more serious problem is economics: a new antibiotic will be approved for use only in the most dire cases, to postpone the development of resistance. This means that, during the patent period, sales will be at best modest.

One promising new antibiotic is Teixobactin, discovered with a new technique for culturing soil bacteria in situ.

Protein pump inhibitors (PPIs): used for various stomach-acid problems, including ulcers

HIV: protease inhibitors moved HIV from a short-term acute illness to a long-term chronic illness. This made drug development profitable again. When AZT (one of the first effective anti-HIV drugs) was first applied in the early 1990's to HIV patients, though, it was an off-label use.

red/white-blood cell drugs:
    filGRAStim/neupogen       makes more neutrophils/other WBCs. Used for cancer/chemo/BMT patients
    erythropoietin          makes more RBCs: kidney disease, cancer, cancer treatment

    exENatide/byetta: this often means the patient can avoid taking insulin. It is offcially for type-2 (non-insulin-dependent) diabetes. From Byetta may also be used for other purposes not listed in this medication guide.

ledipasvir and sofosbuvir: These were approved in 2014 and 2013 respectively. A combination of these two agents is used in the treatment of Hepatitis C; the cure rate for a common strain of the virus is around 95%, and serious side effects are rare. (Gilead Pharmaceuticals decision to price a course of treatment with their Harvoni brand of ledipasvir/sofosbuvir at close to $100,000 has been controversial; see

In 1984, Congress passed the Drug Price Competition and Patent Term Restoration Act, also known as the Hatch-Waxman Act. This allowed generic drug makers to use a patented drug in their own FDA application, so that approval would be in place as soon as the original patent expired (normally after 20 years). However, it also gave pharmaceutical developers a chance at a patent extension for up to five extra years, subject to the following:

The idea here is that if it took a company fifteen years to bring a drug from discovery (and patent) to market, they will get ten years of patent-protected sales to recoup their investment rather than five.

New Drug Rates

Generally, only two to three dozen new pharmaceuticals receive FDA approval each year. In 2012 the number was 39; in 2013 it was 27.

Third-world issues with pharmaceutical patents

Basically, if a country is too poor to afford to treat all its citizens with a new drug, many may die. This has definitely been the case with some HIV drugs. As a result, the Third World has long argued that it should be exempt from pharmaceutical patents.

In the mid-1990's, the WTO basically agreed, and allowed (I'm not sure of the exact terms) third-world countries to manufacture generic equivalents of first-world drugs for use within their own borders only. The manufacturing must be for the government's own use (eg in distributing to its people; the manufacturing can't be a for-profit initiative of a private company in that country). Also, a good-faith attempt must have been made first to negotiate for a reduced-rate license to manufacture the drug, and such profits as are made must be given to the patent holder. In practice, large US pharmaceutical companies often enter into vastly-reduced-price licensing arrangements with third-world companies.

In 2005, the WTO relaxed this rule to allow poorer nations to import generics that would otherwise be covered by a patent. This is sometimes described as a "compulsory license". The manufacturer would not be prosecuted.

This is perhaps the foremost third-world patent issue.

On April 1, 2013, India's Supreme Court rejected an attempt by Novartis to patent Gleevec, known everywhere but in the US as Glivec.

Gleevec exists in two forms: the "alpha" principal form and the "beta" form. The issue before the Indian court was that Novartis was trying to patent the beta form, a small change to the existing alpha form. India added pharmaceutical patents to its law in 2005, but with the proviso that patents for new drugs had to prove significant clinical advantages over alternative formulations. Apparently (this is not 100% clear to me) the alpha form had been patented in India (perhaps the US patent had been recognized?), but that patent had expired.
In Mumbai there is (or was) a large sign reading "Novartis: Making a killing in profits". The slogan remains part of a poster.

See also the above-mentioned article

For a look at the downside of pharmaceutical donations and special pricing, see There is no such thing as "free" vaccines.

Pharmaceutical pricing remains controversial. It was a major sticking point in the 2016 Trans-Pacific Partnership (TPP) negotiations (the TPP now seems dead). The US (and to a lesser extent Europe) subsidizes pharmaceutical development for the rest of the world. Should it? There are plenty of people in the US who cannot afford their prescriptions.

In an editorial in The Economist,, it is claimed

Patents on drugs are in the interests of the sick as well as the industry. Protection should not be weakened.

Patent rights, in other words, are the only reason many new drugs are ever developed.

Writing in The Atlantic (, judge Richard Posner (are judges even allowed to write for The Atlantic?) wrote

So pharmaceuticals are the poster child for the patent system.

But Posner's title was Why There Are Too Many Patents in America.

The US patent on imatinib expired in 2015. Novartis is believed to have paid Sun Pharmaceuticals to delay the introduction of their generic formulation. When the generic was introduced, it was priced at around $140,000/year, versus $146,000 for brand-name Gleevec (see here). The price has since fallen, but not precipitously.

Pharmaceutical companies routinely try to patent minor chemical modifications of existing agents. This sometimes leads to significant clinical differences, but is derisively known as "evergreening". Here are a few stereoisomer examples, with the prefix signifying the isomeric difference in bold:

ARM patents

If you have a smartphone, odds are it uses a RISC ARM processor.

But ARM does not make processors!

Instead, they design them. Companies license the design and then arrange with a chip foundry for the actual manufacture. Licensees often make minor adjustments to the overall layout, though most (the so-called "hard-core" licensees) make no changes to the actual processor cores. A minority (the "soft-core" licensees) do make such changes.

ARM's designs are protected by copyright. It would be very difficult for someone to copy one of ARM's core designs and change it "just enough" so that copyright no longer applied; this would be a major redesign effort.

But ARM's designs are also, sort of as a backup, protected by patent. Even if someone did re-engineer the core so as to avoid copyright infringement, there would still likely be instances of patent infringement. By the time someone designed around all ARM's patents, they might as well be designing an independent processor.

In 2000, ARM sued picoTurbo basically out of existence over processor patents.

In 2012, ARM licensed most of the RISC patent portfolio of MIPS Technologies. This protects ARM from being sued.

Is ARM's business model appropriate?

Patents Generally

The fundamental justification for patents is to encourage advances in technology. Their justification is purely utilitarian; it is difficult to give a deontological rationale for not using someone else's invention.

Ethically, the scope of the utilitarian argument can be worldwide, but national patent laws are generally intended for national benefit. That said, treaties have led to fairly uniform patent laws in the industrialized world, but there are serious concerns in the Third World about certain aspects of patent law.

Also, a part of the patent process is the requirement that the inventor disclose the idea to the world. You cannot protect something with both a patent and trade secrecy. To put it another way, the patent grants you a limited monopoly in "exchange" for publication.

Software Patents

You cannot patent a software application as such; software patents are meant to protect software inventions; that is, particular ideas used. For example, if I am running an online directory service, I might patent hashing as a means of lookup (sadly, I might succeed, even though hashing has been well-known for many decades).

Copyright protects source code and executable code; it has been extended to user interfaces and the Supreme Court is set to rule on the copyrightability of APIs. Copyright protects only the expression and not the underlying idea; when user interfaces are copyrighted, someone else must modify the interface in some ways.

Suppose that in developing your application you come up with a new feature, or particularly innovative algorithm, or innovative application of an existing algorithm, or a novel user-interface gesture (think iPhone), or new way of presenting data, or some other novel feature. You may be able to patent it, as an "invention".  You may even be able to patent a new business method supported by your software. Patents are for certain features that you consider especially innovative.

If you have a patent, then nobody else can use your idea without your permission, even if they discovered and developed it entirely separately, for the lifetime of the patent.

Issues with Software Patents

Why are software patents even controversial? Maybe the basic theory is that, in computer science, almost all ideas are straightforward to implement. This is completely unlike every other field of endeavor, where, after someone comes up with an idea, extensive and uncertain development work is the norm. If you have an idea for a new pharmaceutical, you have to test it in patients. If you have an idea for a new brake rotor, you have to ensure that it doesn't corrode too fast, or get too hot, or have hidden weaknesses. But if you have a new software idea, developing it is pretty much only a matter of time and budget.

Some people argue that patents are fundamentally flawed. That approach, however, leaves us without pharmaceutical patents. Here are some issues that are a little more specific to software:

Complaints about the patent system as a purveyor of undeserved monopolies go back a long time; usually these come from the same sources as attacks on copyright. An implementation issue might be the (rather more mainstream) claim that the patent office is too free in granting patents for weak software ideas. We will get to business-method patents below, but it is not unreasonable to suppose that monopolies on a specific business idea are incompatible with free enterprise.

One example of a fundamental software-patent problem might be that algorithms are can be seen as mathematical facts, not inventions. While novel applications of existing algorithms can be patented, "novel" is very subjective: one developer's breakthrough idea might be a standard software-engineering exercise to another.

Another issue is patents for the application of general, abstract principles to specific problems. This makes a certain amount of sense for physical inventions, because it is not entirely obvious what general engineering idea will work in a given physical context. However, in the world of software this kind of general-principle application is often trivial. There is seldom any question that the method will work, for example; it is as if the developer simply opened a copy of Knuth or CLR and looked up a suitable algorithm in the index. Sometimes it might have taken someone some effort to figure out just which algorithm is most suitable, but should that make the invention patentable?

A related part of the problem, with aspects unique to computing, is that any invention has two parts:
In the physical world, the second often tends to be the much harder component. But in the software world, once you realize you have a need to do something, working out an implementation (at least at the "invention" stage) is often (not always, but often) very straightforward. The developers do standard design work, and look up algorithms in the above-mentioned Knuth and CLR. Sometimes there's an important optimization to be made, because the first draft wasn't fast enough. But that, too, is straightforward, once the limitations of the original approach are clear.

What often seems to happen is that a patent is granted for an invention at a time when the inventor had no clear concept of how to use the idea. Much later, someone needed to get something done, and found a quick obvious solution that turned out to be the earlier patented one; this is the middle issue above. The Eolas case (below), on running applets within a browser, is a possible instance of this. Should early inventors be able to lock out later developers simply because they managed to patent a technique to solve a problem that was completely obvious later to anyone else faced with that same problem? Haven't they really patented the problem?

To put this another way, some inventions are the result of a genuine inventive spark, and some are very obvious. Both kinds of inventions do get patents, although patent examiners try to weed out most of the latter. Most inventions, and probably most patents, arise because someone needed to be able to do something. A plausible test for obviousness of an invention is that, if other "people of ordinary skill in the art" faced the same problem, they would quickly solve it with the given invention. So a nonobvious invention would be one for which other people who worked on it would not immediately arrive at the solution.

Physical-world patents usually take quite a number of real-world "building" steps before they can be realized, and so the total amount of effort tends to make the end result reasonably "non-obvious". However, large numbers of computing patents appear to describe inventions that any competent software engineer would have come up with independently; that is, there is not any real inventive spark. Therefore, when these ideas are patented, they are not generating any social utility. Large numbers of computing patents often seem to represent an approach that any average software engineer would have come up with very quickly.

Obvious in Context

At the risk of being misleading about the term "obvious" as used in patent law, I will call this the "obvious-in-context" problem. Much of what passes for "invention", at least in terms of application of known algorithms, is really just competent software engineering. We might define when an invention is obvious in context as follows:

If a software engineer of ordinary skill, familiar with the general area of programming, and familiar with common algorithms, and who was presented with the problem in a reasonably concrete setting, would readily come up with the solution in question.

This is quite different from readily being able to come up with an implementation, given the solution. That might take significant coding, but might involve little if any inventive leap. The idea here is that the patented idea would be obvious to anyone presented with the problem it solves.

This not, however, quite the same as the patent standard of "obviousness" under current law, in which discovering the applicability of an idea to a particular domain can be part of the novelty of an invention. For example, consider the use of XOR for drawing easily erasable objects on a video screen. We can take either of the following two approaches:

Perhaps a more serious issue is that the problem to be solved may have arisen long after the patent. That is, an engineer, when faced with a certain problem in 2015, may find an "obvious" solution that infringes on a patent filed in 2010. The engineer may feel the solution would have been equally obvious in 2010. But perhaps the problem in question couldn't have been formulated in quite the same way in 2010, or perhaps the problem itself suggests some element of the 2010 patent invention. The usual patent-office standard for obviousness means that the entire solution would have been obvious as of 2010.

An example here might be the Eolas applet patent. Web applets with embedded controls became obvious after browsers became ubiquitous, but Eolas was granted its patent in the early days of browsers, when embedded controls might have seemed much more novel. (That said, Eolas eventually lost its patent case because of evidence of prior art.)

A hypothetical solution to a problem of physical engineering is not the end of the story; it remains to be seen whether the solution works in practice. But for software engineering, that is not a concern: unless the engineer has missed something, the solution will work.

There are several specific facets here:

Another way to describe this situation is to say that many patented software solutions are ordinary, obvious applications of basic principles of software engineering.

Patent-law obviousness

When a patent case comes to trial, the judge will often take the position that if the patented idea really were "obvious", then someone else would have patented it first! Or at least published it first. This (codified into the original set of rules for interpreting "obviousness") can be an enormous hurdle for the patent defendant. Nonetheless, for software anyway the obvious-in-context rule suggests that an "invention" can be obvious, but only of interest when other infrastructure had reached a certain level.

Another judicial hurdle is that, when faced with a claim that a patented invention is obvious, the judge will argue that it is only obvious in retrospect, and the defendants must prove it was obvious at the time of invention. This is impossibly hard, unless there is actual published work prior to the time of the invention that describes the invention. Such a publication would then be considered prior art.

If prior art is published, it can invalidate a patent. However, if it was used privately, those users can continue to use their idea without paying royalties to the owner of the patent, but the patent may still stand. The patent can be challenged on the grounds of not being novel, but this is harder.

As we consider patents below, it will often be claimed that the innovation is "obvious". In fact, there is sometimes (maybe even often!) evidence that the idea was so obvious that prior publication (required to invalidate a patent based on "prior art") would have been rejected by any reputable publisher.

In the first facet in the list above, the obvious-in-context problem boils down to a general algorithm being applied to solve a specific problem. The Flook case (below) supposedly made many such inventions non-patentable. However, this isn't always what happens in practice, where the decision comes down to the question of just how much innovation was needed in applying the algorithm.

While one can argue that a good patent system can have a very positive effect on business and on investment in innovation, it seems clear that a bad patent system can have a very negative effect. Good patents reward innovation but bad patents actually punish it.

This brings us to a second patent issue that is at least somewhat software-specific: as a practical matter, you cannot tell what are the boundaries of a patent claim. This is slightly due to obfuscatory language, but the larger problem is the flexibility plaintiffs have in "claims construction", that is, in deciding what the patent covers. The E-data system, below, is an example of this.

A third issue with software patents is functional claims: patents for a certain transformation, regardless of how it is implemented. One example is the Ultramercial patent on displaying ads within a video stream, as is done by Hulu. Another might be the Amazon "one-click" patent. No pharmaceutical manufacturer, on the other hand, would even dream of filing a patent for "any chemical that cures the common cold".

The smartphone wars may or may not be an example of a positive use of the patent system. If Apple invented some breakthrough phone features for the iPhone, and patented them, then Google should perhaps not be allowed simply to copy them in Android-based smartphones. But exactly what happened here is less than clear; we will revisit it below. Apple did not, as it turns out, invent the "smart phone". Nor did they invent the multi-touch interface. They did not even invent the graphical interface; see

One of the central problems in the history of "patent wars" in any field is the exceptional expense of patent litigation; another is that the law puts the burden of proof on the defendant to prove a patent is invalid; patents are presumed valid until proven otherwise. Proving a patent's invalidity is not easy; realistically, this involves presenting "prior art". Direct challenge on the grounds of obviousness is particularly difficult. An added issue is that the party suing may be a patent-holding corporation that cannot be effectively countersued. Judges are often sympathetic to the "lonely inventor" whose idea is being "stolen" by a "big corporation". Finally, court costs are rarely awarded to those who win patent-infringement cases, which sometimes seems to have the effect of encouraging frivolous suits.

KSR and Obviousness

As stated earlier, the "obvious-in-context" standard is something like this:

If a software engineer of ordinary skill, familiar with the general area of programming, and familiar with common algorithms, and who was presented with the problem in a reasonably concrete setting, would readily come up with the solution in question.

The KSR v Teleflex case from 2007 does present one practical way of establishing obviousness: an invention is obvious if it is a straightforward combination of two (or sometimes more) prior-art ideas. This is related to the "obvious-in-context" idea because if a solution is obvious given the problem, then the solution is probably a combination of prior-art ideas (that is, well-known ideas), and therefore a KSR-type argument can often be made. Still, just because an invention is a combination of prior-art ideas does not always mean it is an obvious combination; there has to be some "motivation to combine". Sometimes there are inventions that are quite simple to implement, but which are not obvious.

Here's an example that might go either way: On the one hand, "interactive pop-up books" is a combination of traditional pop-up books with standard ways of making content interactive by adding LEDs and buttons and chips and absolutely-standard stuff like that. On the other hand, bringing computer interactivity to children's pop-up books might be considered to be an innovative step.

In principle, the KSR decision rules out patents that are simple combinations of prior-art ideas, but doesn't apply to single-concept inventions. What about the XOR-mouse patent? Is this a single concept? Or a combination, of the XOR algorithm and the bitmapped video system?

List of software-patent issues

Abstraction: Often patents cover the application of general known algorithms to special cases. This seems intrinsically obvious.

Abstraction: Often patents cover "abstract ideas", which are not supposed to be patentable.

Obvious-in-context: Many software innovations are obvious, but it takes a specific context to make it worth the trouble to implement them.

Functionality: Some software patents appear to cover a specific function, not a way of implementing that function.

Using a Computer: it is inappropriate to patent an existing process that is now being implemented on a computer.

Claims Inflation: patent "trolls" often rely on the broad interpretation of a patent, so as to cover an idea that arose years later. At the time the patent was filed, the later idea was nowhere in evidence. See E-data.

Ambiguity: Claims are often written in very obfuscatory language. Even clearly worded software claims often have very poorly defined "boundaries"; ie they admit many possible different interpretations.

Aeroplane Control

A classic "broad" patent is the Wright brothers patent on "wing-warping" to control flight (to the Wrights, the wings were the "aero-planes", planing the air, the rest of it was the "flying machine"). The Wright brothers actually twisted the whole biplane structure -- using cables -- to bank in a turn. This later led to the development by others of ailerons, which achieve the same effect but which mechanically are entirely different. A court ruled the Wright patent still applied: what mattered was the concept of adjusting wing angles to tilt the craft.

This is a classic "broad patent" for a major new innovation. The courts have traditionally recognized "broad" patents, but somehow in software this distinction is often lost.

Some have argued that the Wright brothers tried to claim a functional patent here: a patent on any method for turning an aircraft. This is not entirely correct: the Wright claim was really for a specific set of movements of the wing and of the rudder.

There has also been major patent litigation about sewing machines, agricultural machinery, 19th-century telephones, and electric power. See

Supreme Court cases limit the word "process" in USC Title 35, Chapter 10, §101:

Inventions Patentable: Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.

Gottschalk v Benson, 1972

This case continues to be debated, although the Supreme Court has recently reaffirmed it. (Gottschalk is the name of the patent commissioner; the one-word name for the case is Benson.) The court stated that "the patent would wholly pre-empt the mathematical formula and in practical effect would be a patent on the algorithm itself". But that is not entirely clear. The algorithm was for converting so-called binary-coded decimal (in which decimal digits are represented in sequence each by four binary digits, so 157 = 0001 0101 0111) to true binary. The patent was for the use of this algorithm on a computer, not universally (though the algorithm makes no sense except on a computer).

In the decision, the court cited an earlier decision: "[w]hile a scientific truth, or the mathematical expression of it, is not a patentable invention, a novel and useful structure created with the aid of knowledge of scientific truth may be", and "an idea of itself is not patentable." The court also said,

Here the "process" claim is so abstract and sweeping as to cover both known and unknown uses of the BCD to pure binary conversion. The end use may (1) vary from the operation of a train to verification of drivers' licenses to researching the law books for precedents and (2) be performed through any existing machinery or future-devised machinery or without any apparatus.

Applications of computer algorithms may indeed be general. In retrospect, should this be an issue? The end-uses above are all for software using the algorithm; and anyone could always use a different algorithm.

One interpretation of the decision quote above is that the Supreme Court is willing to allow patents on algorithms that are an integral part of an invention, but not on algorithms that are in effect simply library routines. The Benson algorithm was clearly in the latter category; if it had been closely tied to a particular invention -- that is, if there had been a claim limitation -- perhaps it would have been patentable. Does that make any sense?

One school of thought is that a "process" is not patentable unless there is a physical transformation involved, or unless the new invention is necessary in carrying out the process; this is sometimes referred to as the "Machine or Transformation" rule. The Supreme Court's Benson ruling was narrower than that, however. Nonetheless, the court did say

transformation and reduction of an article 'to a different state or thing' is the clue to the patentability of a process claim that does not include particular machines. [emphasis by pld]

At one time, I might have put the emphasis differently: the clue. However, the Supreme court backed away from that position in the Bilsky case (below).

The Benson decision closes with an observation that lack of patentability doesn't seem to have hurt the 1972 software industry any, and a call to Congress to figure it out (which may have had its origins in an amicus curiae brief filed by IBM, which was at the time against software patents):
If these programs are to be patentable,  considerable problems are raised which only committees of Congress can manage, for broad powers of investigation are needed, including hearings which canvass the wide variety of views which those operating in this field entertain. The technological problems tendered in the many briefs before us indicate to us that considered action by the Congress is needed.

One possible way to understand Benson is that the patent claim limitation -- the application of the invention for which the patent was claimed -- was overly broad. Benson claimed a patent on any use of the algorithm on a computer.

In the 1980 Supreme Court decision Diamond v Chakrabarty, about the patentability of genetically modified organisms, the following line from a 1952 Congress report on patent law is cited

    anything under the sun that is made by man [is patentable]

How does this square with Benson?

The Benson decision is frequently cited as evidence that the Supreme Court did not intend to allow software patents; the language of the decision -- that the algorithm was a fundamental mathematical fact -- can be read as supporting this. Others use Benson as an example of a case that, with the benefit of hindsight, seems more and more strongly to have been wrongly decided; that in patenting an algorithm on a computer, one is not patenting a mathematical idea. Or, at the very least, Congress should simply heed the Court's call to address this issue directly. It doesn't help any that the court seems to have misunderstood the concept of widely varying applications.

In the Diehr case, below, the Supreme Court seems to have allowed the patenting of a much simpler algorithm (really not an algorithm at all) in the context of a specific application to a method of manufacturing rubber. Is the only issue with the Benson patent the fact that it was so general? Aren't most software ideas very general?

Parker v Flook, 1978

In this case, the Supreme Court dealt with an invention that used a computer to analyze temperature fluctuations within a catalytic converter, apply a smoothing algorithm to discount transient fluctuations, and raise an alarm when the averaged temperature was outside of preset limits (the algorithm apparently also had something to do with updating those limits).

The only novel feature of the method is a mathematical formula. In Gottschalk v. Benson, we held that the discovery of a novel and useful mathematical formula may not be patented. The question in this case is whether the identification of a limited category of useful, though conventional, post-solution applications of such a formula makes respondent's method eligible for patent protection.

The court ruled that the patent had to be judged as if the algorithm at stake were prior art; that is, the algorithm cannot be counted as part of the innovation.

Respondent's process is unpatentable under 101, not because it contains a mathematical algorithm as one component, but because once that algorithm is assumed to be within the prior art, the application, considered as a whole, contains no patentable invention. Even though a phenomenon of nature or mathematical formula may be well known, an inventive application of the principle may be patented. Conversely, the discovery of such a phenomenon cannot support a patent unless there is some other inventive concept in its application. [emphasis added -- pld]

One difference between Flook and Benson was that in the Benson case there was essentially no claim limitation, while in Flook the claim was limited to a specific application of temperature control.

There has been some criticism of Flook because the Supreme Court was vague as to the distinction between §101, on patent eligibility itself, and §103, on restrictions due to prior art and obviousness.Such a low rate would have caused potential patent plaintiffs to think twice about filing a lawsuit using a weak patent.

Flook is arguably an important case, but alas often seems overlooked in the real world. Consider, for example, a patent for the application of a standard algorithm to a particular problem:
Both of these patents were granted. Arguably, however, both are trivial applications of standard algorithms. In fact, more often than not patents that use general, well-known algorithms don't seem to be rejected on Flook grounds, even when there is essentially no inventiveness in the application of the algorithm.

Diamond v Diehr, 1981

On the face of it, the Supreme Court ruled here that just because an invention incorporates a software component, this does not make the invention ineligible for patent. The Diehr invention used a thermocouple to monitor the temperature in a mold of curing rubber, connected to a computer that read the raw temperature values and did some computation to determine when the rubber was finished. While the majority held that the invention was just patentable, Justice Stevens (of Sony v Universal) wrote in dissent of three problems he had with the Diehr patent [division into paragraphs added]:

First, there is not a word in the patent application that suggests that there is anything unusual about the temperature-reading devices used in this process -- or indeed that any particular species of temperature-reading device should be used in it.

Second, since devices for constantly measuring actual temperatures-on a back porch, for example-have been familiar articles for quite some time, I find it difficult to believe that a patent application filed in 1975 was premised on the notion that a process of constantly measuring the actual temperature had just been discovered.

Finally, the [PTO] expressly found that the only difference between the conventional methods of operating a molding press and that claimed in [the] application rests in those steps of the claims which relate to the calculation incident to the solution of the mathematical problem or formula used to control the mold heater and the automatic opening of the press.

In other words, Stevens felt that the Diehr invention should not be patentable, because the only novel feature was the use of the computer, and algorithms should not be patentable.

The Federal Circuit

Diehr was decided in 1981. In 1982, Congress created a special appeals court for patent cases, the Federal Circuit; all patent cases appealed from the federal district courts then went to the Federal circuit. Patent cases are technically complex, and pose a problem for judges. The theory was that the Federal Circuit would thus gain expertise in patent cases, and would be better able to rule consistently.

The Supreme Court then got out of the patent-case business for a long while, presumably in order for the Federal Circuit to develop its own legal theories.

There are still arguments in favor of this single-patent-circuit idea: patent decisions in different circuits were often significantly different, and "general-purpose" district-court judges typically did not have the technical background to understand some patent issues. However, there is also a theory that the Federal Circuit plan backfired massively. See, especially the article's title.

Overall, the Federal Circuit appears to have been from the beginning very friendly to patent claims, and unwilling to address the broader interests of the rest of society in having patents invalidated. Perhaps we all benefit from Apple's investment in innovation that might not have been made without the patent system to protect it. But for every Apple (and there is really only one), there are thousands of petty patent cases that in effect reduce the incentive to innovate. If an innovative solution will cost you tens of millions in unforeseen patent litigation, it's best not to bother.

Here is a quote from the rogue-appeals-court article above:

In the two decades before the creation of the Federal Circuit, appeals courts sided with patent holders in only 20 to 40 percent of cases. ... In contrast, the Federal Circuit sided with patent holders more than 80 percent of the time during its first year on the bench.

It was the Federal Circuit that developed the original patent test for "obviousness": there had to be documented evidence in the prior art that suggested that someone was aware that the combination of new features in the new invention might be desirable. This was known as the "teaching-suggestion-motivation" test. It led to the patenting of many software ideas that might be so obvious that nobody would have bothered to write them down.

The Supreme Court's Benson, Flook and Diehr patents might have been read as banning software patents. It was largely the Federal Circuit (with the complicity of the Patent Office) that undermined these precedents. And it was the Federal Circuit that ruled, in the State Street Bank case (below), that business methods were patentable.

By making the Federal Circuit hear only patent cases, the judges focused solely on patent issues. The interests of the public were nowhere represented.

Reading original patent descriptions

You can look them up online. To use the patent office files, go to the search-by-patent-number page:

Google can also be used to search for patents at; their format is often better.

The legal basis of a patent lies not in the description, or the abstract, or the illustrations, but in the claims. See
for more information.

PTO (Patent & Trademark Office) problems:

ignorance is no defense: "submarine" patents
The entire process is secret: you can be making good-faith effort to be noninfringing and get hit with a huge verdict.

Infringement is considered willful if you had advance notice of infringing. Your belief that the patent was invalid is NOT generally a defense (although some judges have recently been allowing this). Damages automatically triple.

Maybe the central problem is that at the time of patent application, there is nobody representing the interests of the public. (Later on there is now a period when the new patent application is public, and there is now a large community devoted to objecting to such patents at that stage.)

A good example of the failure of the PTO is the issue of granting a patent (11047359) to a perpetual motion machine. Such a machine cannot exist, and so cannot be "useful". Yet this patent was granted. See for the history, including the patent examiner's initial denial and a discussion of how the inventor wore down the examiner's resistance.

For another tale, this time about a perpetual-motion patent that was rejected, but maybe not for the right reasons, see

Three Groups of Stakeholders:

Where do we fit in?

Patent-holding companies are sometimes known as "trolls", as in "the troll under the bridge, demanding tolls". They are often seen as parasitic: they do not, for example, actually produce anything. Is this a fair characterization? They do, after all, create a market for small inventors.

How large corporations manage:

Small inventors:

Many times, when patent reform is proposed it is the small-inventors community that is most up in arms. Should we care?

Legal advantages of small inventor: somewhat diminished with rise in legal fees & increased ambiguity of patent claims

But small inventors can still sell to patent-holding companies.

Legal situation of large corporations:

Examples of software patents

Here's an interesting list of some software patents, supposedly "random" though some particularly famous ones are sprinkled in:

Another list can be found at:

In the examples here, try to keep separate the idea of the invention from how you might implement it.

CPU Stack Pointer register

Patent 3794980, to George Cogar, 1971. Supposedly, though I can't find the source, this was purchased by DEC [?] and eventually placed in the public domain. Or not enforced. Or, possibly, it was in fact licensed by Intel/Motorola/everyone_else.

ATT setuid-bit patent

Patent 4135240, invented by Dennis Ritchie, applied for in 1972 and later placed by AT&T in the public domain. The Supreme Court had just ruled in Benson that software couldn't be patented; ATT apparently got the patent because they claimed it was hardware. This patent was dedicated to the public domain in 1979. This patent is certainly a deep idea: if a certain bit is set in the filesystem information node for a file (not in the file itself), then when the file is executed, it runs with the privileges of its owner and not the user. Before then (and after; look up why Windows has all those svchost processes), there were complex ad-hoc methods for running selected programs with elevated (or alternative) privileges.

British Telecom Hyperlink patent

Patent 4873662. BT filed in 1976; the patent was granted in 1989. In 2000 they attempted to sue a large number of websites. The patent itself apparently refers to a system in which multiple remote users can access "blocks" of data on a central computer.

Here, the arguable problem is that BT's claims are much too broad and vague.

Altavista patents on web search

Altavista did have several such patents on basic web "crawling". Most were predated by the "archie" archive-searcher for ftp sites, developed by Alan Emtage in 1989. Altavista was purchased by CMGI, which then tried to demand licensing fees for its patents.

RSA encryption

Patent 4405829: Here is a short description

Choose primes p and q.
Reveal n=pq publicly.
Find e and d so xed = x mod n (this is not hard) Common values for e: 3, 5, 17, 257, 65537. Note (xe)d = xed.
(n,e) is the public key, d is the private key. A plaintext p is encrypted as c = pe mod n; it is decrypted as cd mod n = ped mod n = p. The plaintext p is most likely a key for a conventional cipher.

RSA uses standard high-precision arithmetic in its calculations; the underlying number theory has been well-known for centuries. The patent was for the APPLICATION of these standard methods to encryption. That application was certainly novel.

The RSA patents finally expired, after much controversy. Part of that controversy was self-induced; the patent owner at one point appeared to grant a general license for noncommercial use, but then backed away from that.

The RSA idea is not "obvious-in-context". In fact, RSA is, for many people, the classic Good Example for why software patents are needed.


Compton filed a patent in 1989 on multimedia, despite Apple Hypercard in ~1987.

Steir patent

Patent 5,060,171 on artificially adding hair to a person's image [Garfinkel article; obvious-in-context; Flook standard?]


This is a voice-compression patent for telephony use. The normal data rate for voice is 64 Kbps; G.729 reduces this to 8 Kbps. Given the idea of such compression, figuring out an implementation is definitely nontrivial, and figuring out an implementation usable in real-time for simple CPUs is harder still.

But eventually open-source alternatives to G.729 appeared. The problem was that G.729 was the standard; alternatives were usually not recognized by at least one step along the path.

Patents on G.729 are now expired (as of 2017)


This is the canonical example of how a vague patent description can lead to "claims expansion". The original patent was for an in-store kiosk at which a customer could select a number of music tracks to be written on the spot to a cassette tape. But E-Data argued that it also applied to home music sales, like iTunes.

Eolas applet patent

About a way for running "applets" in a browser window, with embedded controls. See below. A good candidate example for "obvious in context"; the idea is obvious to anyone thinking of the browser as a universal interface. (That the implementation is also obvious is just a given.)

NTP v RIM: the blackberry patent. See below.

Spreadsheet natural-order recalculation

Early spreadsheets restricted cells to referring to cells in earlier rows, or in earlier columns. Natural-order recalculation means that any cell dependencies are allowed, as long as the dependencies are not circular. The patent was filed in 1970; the topological-sort algorithm for implementing it was published in 1963. The spreadsheet was not invented until 1979. Once the spreadsheet was invented, natural-order recalculation becomes obvious.

Using XOR to write to video memory

You use XOR once to display the content, and a second time to erase it. Was this well-known enough in 1976 to be obvious? Again this is a candidate for "obvious in context"; anyone needing to write to and then restore a memory buffer (eg a video buffer) would naturally think of using XOR. But also note that XOR is the implementation, not the "idea".

Lempel-Ziv / LZW compression, patents 4464650 [?], 4558302

Lempel and Ziv published their compression scheme in 1978; Terry Welch published an improvement in 1984. CompuServe used this compression mechanism for the GIF file format, introduced in 1987. A year later, they noticed that the algorithm was patented, and that the patent was currently held by Unisys. Allegedly, Unisys told Compuserve at that time that they would not need to pay royalties.

In 1999, Unisys demanded that some noncommercial websites pay a $5000 fee for hosting files in the gif format (an interesting strategy, given that the sites were not themselves rendering the images). Some commercial sites were asked for even more. It's not clear whether anyone paid it; most affected sites rapidly switched to .jpeg or .png. Some observers were especially offended by the fact that Unisys allowed the use of the GIF format as a free standard until it became well-established, and then demanded fees.

There's some question as to whether this was the only or even the dominant reason for the shift to PNG format; the latter does offer more features (especially alpha and gamma) than GIF, and is a lossless format unlike JPEG.

Note that LZW compression is not "obvious-in-context"; you can realize you need compression and still work very hard to develop LZW.

Newer compression techniques are even less obvious. These would include the patents on the MP3 sound-compression strategy, and the G.729 voice-compression algorithm.

Bandwidth-sharing patents

There are a number of patents for deciding how to share bandwidth among a set of customers (eg of an ISP, or guests at a hotel). They all depend on the deficit-round-robin implementation of Weighted Fair Queuing. Typically these take the form of a strategy for sharing unreserved bandwidth in support of a specific business goal.

MP3 patents and lawsuits

The MP3 idea was not obvious, and remains fairly complex. Alcatel-Lucent v Microsoft: Alcatel-Lucent won $1.5 billion in an infringement suit about mp3 decoders Feb 22, 2007

MS countersued for other patents

The judge eventually set aside the damages, and the appellate court agreed.

Aug 6, 2007: MS won new trial

MS then sued A-L for other patents.

check out (Thompson) Royalty Rates: basic mp3 decoder: $0.75/unit

MP3 was published in 1991. Did all US mp3 patents expire in 2011? Original holder: Thompson Consumer Electronics & Fraunhofer Institute. These still hold the "core" mp3 patents.

MP3 Patent claimants:

To date, (some) patent holders have announced that no action will be taken against open-source decoders.

The mp3 compression algorithm is admittedly a deep idea. Part of it involves the use of wave decomposition to store the information more efficiently; part of it involves "psychoacoustics" to identify parts of a sound file that are "unhearable" and so can be deleted.

MS has patented the FAT disk format. Here is a summary of the events from the Public Patent Foundation, which had challenged that patent:

In April 2004, the Public Patent Foundation filed a formal request with the United States Patent and Trademark Office to revoke Microsoft Corporation's patent on the FAT File System, touted by Microsoft as being "the ubiquitous format used for interchange of media between computers, and, since the advent of inexpensive, removable flash memory, also between digital devices." In its filing, PUBPAT submitted previously unseen prior art showing the patent, which issued in November 1996 and is not otherwise due to expire until 2013, was obvious and, as such, should have never been granted. The PTO granted PUBPAT's request in June 2004 and provisionally rejected the patent in September 2004. In response to the rejection, Microsoft made amendments to the patent's claims. The PTO proceeded to allow the amended form of the patent. PUBPAT had no right to appeal that decision.

The fact that PubPat can't appeal or even respond to the decision is a serious procedural issue. If you are sued by Microsoft for using the FAT format, you will be allowed to respond to Microsoft's post-2004 amendments to the patent claims. But you will probably not.

Microsoft is now actively licensing the format to flash-card manufacturers. Why not switch to the free ext2 filesystem? Because that is impossible if the device in question has embedded software to read the FAT format.

It is unclear which patents Microsoft is most interested in pushing. While there are legitimate reasons for thinking that FAT itself might not be patentable, Microsoft later introduced a strategy for supporting both long filenames and also "8.3" filenames (8 char name, 3 char extension). The patent that covers the dual-name strategy is less vulnerable to "obviousness" or "prior art" challenges.

Microsoft sued the TomTom people (, because TomToms, although they are based on Linux, use the FAT filesystem internally. The suit was settled two months later, after TomTom agreed to switch to the ext2 filesystem.

FAT is a good example of an idea that is obvious to anyone presented with the problem of how to create a simple filesystem suitable for small devices. However, there is the added problem of it's being standard (although an unofficial one); you can't just implement a workaround.

Photoshop and patents

It might seem reasonable that such a complex program as Adobe Photoshop has been effectively protected by patents. However, I can find no evidence that this is the case. I did find the following:

Macromedia has sued Adobe for patent infringement over some Photoshop options (

The leading open-source alternative to Photoshop, gimp, may have had more trouble with Microsoft patents than Adobe's. In general, there are lots of notes out there from gimp developers discussing their patent woes. (I lost my specific reference, though.) Gimp has had to struggle with the LZW patent on the gif format (US expiration 2003, Europe 2004), and the possible jpeg patent.

Patent infringement suits

Note that we do not generally have much access to the judge's reasoning in these cases. There may be a written opinion, but often not, and in any event we generally don't have ready access to many of the documents offered in evidence.

(Some of the documents, though not all, are available from the Law Library.)

Heckel Paper

(In the link above, correct the spelling of "pdf")

Paul Heckel wrote Debunking the Software Patent Myths for the June 1992 Communications of the ACM. Heckel approaches software patents as a small inventor who has sued Apple over Hypercard and is in general very strongly pro-patent. He was actually faced with huge software-development debts and no future market because Apple had reproduced his idea. He makes several claims about some well-known early software patents.

4197590, January 1976

Method for dynamically viewing image elements stored in a random access memory array

The use of XOR here (using XOR to draw and also to erase) was mathematically well-known. Is this an innovative application, or a straightforward one?

Heckel calls the patent "largely the invention of the frame buffer". However, the concept of a framebuffer, or memory-mapped display (in which the video screen is memory-mapped into a location in RAM, so that writing to the screen is as simple as writing to memory) was definitely prior art; the patent dates from 1976 but there was an article on the idea by Michael Noll in the Communications of the ACM (not a research journal) in 1971. Evans and Sutherland began selling framebuffer hardware in 1974 (for ~$50,000!). The problem with frame buffers was that, in the 1970's, RAM was extremely expensive! Framebuffers were an idea that wasn't so much as invented as waited for the price to come down.

The patent application claims that they invent a way for a small (eg 416x312) graphics window to allow viewing of a larger 2048x2048 virtual display, and panning across. That is a significant invention. However, Cadtrak attempted to enforce the XOR-cursor portion of the patent alone.

From the patent abstract:

A computer graphics display system including random access raster memory for storing data to be displayed, a raster memory control unit for writing data into the raster memory, a video control unit for causing such information to be displayed on a CRT display screen....An XOR feature allows a selective erase that restores lines crossing or concurrent with erased lines. The XOR feature permits part of the drawing to be moved or "dragged" into place without erasing other parts of the drawing.

It seems to me that a major claim of this patent, and the only relevant claim past the 1984 introduction of the Macintosh, is the XOR feature.

It is legitimate to try to enforce just a single claim; if the Patent Office thinks any of your claims are too obvious, they are supposed to say so. However, claiming that the XOR is nonobvious makes at least some sense in the context of the original patent; the idea is much more obvious as a standalone operation.

For the XOR patent, the context was the development of memory-mapped video. At that point, XOR became the "obvious" way to update the screen so as to allow a very fast and easy undo. But without practical experience with memory-mapped video, the idea simply seems obscure.

4398249, August 1970
This is the natural-order-recalculation patent, 4398249; commonly used in spreadsheets. This was an invention of Rene K. Pardo and Remy Landau in 1970.

It is a classic example of applying a well-known algorithm to a specific case.

The idea is to take a set of formulas (spreadsheet cells, for example) that have dependencies: some cells depend on other cells in that they use the value of the other cell when computing their own value. We want to find a way to evaluate all the cells correctly, unless there is an actual circularity.

Cell A depends on B if A needs B's value in its formula
Rule: Before calculating A, calculate all cells A depends on.
Normally this is done iteratively: first evaluate all cells that depend on nothing, then those cells that depend on the ones just done, etc etc.

More formally:
Let Known_Cells be the set of all cells we can calculate initially, ie that do not depend on any other cells.
     search the spreadsheet for cells which depend only on Known_Cells.
     Calculate their value, and add them to Known_Cells
until there are no new Known_Cells

The algorithm is called "topological sort"; published in the CS literature in 1963.
Heckel claims that what Pardo and Landau really invented was the spreadsheet, although (apparently) in primordial form not recognizable by later spreadsheet pioneers such as Dan Bricklin (VisiCalc, first released in 1979); Bricklin is generally recognized as the inventor of the spreadsheet. Heckel does not name any software product Pardo and Landau produced. There was indeed an important legal case regarding the patentability of the idea; initially their request was denied by the PTO as an "algorithm. It is now considered to be an "early" software patent.

Spreadsheets were a brilliant idea (Dan Bricklin, VisiCalc?). How significant, as an invention, is natural-order recalculation? Was it a major insight? Or did someone finally pick up their copy of Knuth and look for an appropriate algorithm?

Bricklin, by the way, did try to patent VisiCalc, but was unable to, because in 1979 (two years before Diehr) he was unable to. VisiCalc was a software application. See There might have been some patentable general principle -- after all, Pardo and Landau managed to get their patent in 1970 -- but officially algorithms could not be patented since Benson in 1972. Pardo and Landau just managed to get lucky.

Pardo and Landau sold the patent to the company Refac, which Heckel characterizes as "a white knight in the fight against the patent pirates". However, note the following line from the appellate court in Refac v Lotus, at

The district court did not abuse its discretion in holding that the '249 patent is unenforceable on the ground of inequitable conduct. [emphasis added]

Apparently Pardo and Landau were told by the USPTO that their patent was not fully disclosed; specifically

[T]he disclosure is not deemed of the level which would enable one with ordinary skill in the art to make and use the invention. The flow diagram which applicants represent in figure 2 is not of the detailed level which a programmer would need in order to write a program from.

They presented three supposedly disinterested witnesses who each submitted an affidavit that, yes, they could implement the desired program from the patent description. However, one of the three, Robert F Bullen, had a previous business relationship with Landau, that neither of them disclosed.

So much for "white knights".

There remains the broader concern that Pardo and Landau did not in fact appear recognize that they had a "spreadsheet", and that Dan Bricklin was the first to grasp that concept, and that what Pardo and Landau did have was just the idea of using topological sort when evaluating a large number of expressions with dependencies.

Here's the abstract of Pardo & Landau's patent, slightly edited, for compiling a set of formulas when some formulas may depend on the values of other formulas:

A process and apparatus (compiler program) carried out on a digital computer for converting a source program into an object program. The source program [consisting of a set of formulas] is entered into a first storage area of a computer. ... The compiler program then examines each formula to determine whether it has been defined [that is, whether it has no dependencies on as-yet-undefined formulas]. If the first formula has been defined, it is removed from the first storage area; placed in a second storage area; and marked as being defined. if the first formula examined has not been defined, it is retained in the first storage area. The compiler program repeats this process for each formula in the first storage area. After the formulas in the first storage area have been examined, the compiler program determines whether any formulas have been added to the second storage area. If so, the compiler program repeats the examination of the individual formulas still retained in the first storage area. The compiler program repeats the process until all formulas have been defined; added to the second storage area; and marked as being defined. The content of the second storage area is an object program which is executable by the computer.

Heckel's own patent was 4736308. It was described as covering the display of two or more strings together on the screen (apparently scrolling with two or more subwindows). The argument was made that if scrolling was prior art, and subwindows were prior art, how could a trivial combination of them be patentable? Heckel's claim is that his patent actually covers a Hypercard-like product (hypertext plus button-activated scripts), though this is not clear from the patent abstract.

As in other software patent disputes, Heckel's idea to some extent fits the model of the application of general software engineering principles to solve a particular problem. However, Heckel's patent also has an element of how to frame the programming environment for the user; this would appear to be a genuine user-interface innovation. (Though note that much of the innovation was done by Xerox in the development of the Smalltalk environment.)

Eolas v Microsoft

Patent 5838906: Distributed hypermedia method for automatically invoking external application providing interaction and display of embedded objects within a hypermedia document, filed in October 1994.

The alleged invention is about a way for running "applets" in a browser window. Is this really an "invention" at all?

Back in 1992, browsers just displayed images and static text. However, it was clear by then that applications could display content in windows created and "owned" by other applications; for example, MS Word could at that time display an Excel subwindow.

What was not foreseen was that browsers would become universal viewers of just about everything. Once they did, though, the Eolas "invention" -- displaying a certain kind of content in a browser -- does become obvious. Should this make the idea not patentable? Eolas finally did lose their patent, but not for this reason.

Microsoft v Eolas (+ Univ of California, as part of UCSF):
MS lost this case in 2004; was ordered to pay $521 million.  They then won the right to a new trial, but settled in 2007 before the beginning of the new trial. There are suggestions on the web that the settlement amount was in the range of $100 million. These may be pure speculation, but the odds are they settled for less than the original verdict.

Between the end of the first trial and the settlement, a PTO re-examination upheld the patent. However, the examiners discovered that a crucial feature of the Eolas patent that was not prior art was that the applet began running immediately, with no user input (like YouTube videos do today). Microsoft modified Active-X for a while so that it did not start automatically; users had to press "run". This "workaround" possibly gave Microsoft an edge in the negotiations.

The patent covers

a system allowing a user of a browser program ... to access and execute an embedded program object [or small computer program, often referred to as "applets" or "plug-ins"]

Patent filed 1994, granted 1998, USPTO review 2004, upheld 2005


"Viola" prior art: may or may not be relevant: see

The Viola browser project was found by the district court to have been "abandoned", but the circuit court found that Viola version 1.0 was "abandoned" only in the sense that it was replaced with version 2.0.

Part of the technical issue was about the meaning of the term "executable application". The court allegedly gave this broad meaning. Microsoft claimed it only meant "standalone applications".

Eolas was started by Michael Doyle, faculty member of UCSF. UC Berkeley is apparently the owner of the patent.

Doyle certainly had some early ideas, that came before Java applets.

Had the Eolas patent been more aggressively enforced (and been more enforceable), it would have been a huge impediment to the world-wide web.

October 2009: Eolas has launched a major lawsuit against just about everyone: press release. The list of defendants includes lots of sites that use web-embedded objects: amazon, apple, eBay, and youtube.

View 1: Eolas developed some early web-embedding technology, and got really lucky
View 2: Eolas patented Microsoft's 1990 Object Linking and Embedding idea, and got a 20-year license to fleece us
View 3: Eolas might control a "critical" piece of the web standards.

Consider the actual Eolas patent claim again. They have allegedly patented the idea that an applet could be displayed automatically within a browser window. In a sense, they have patented the idea that browsers can display certain kinds of content.

(Actually, their claim oscillates between the idea of displaying an applet, and the idea of displaying it automatically. This is a tricky issue.)

Who had the idea that browsers could display any content, and thus that that was all you really needed? One version of VMware used a browser plugin to view your virtual machine's console!

Maybe the deepest problem with the Eolas patent is that the whole idea of embedding one window in another is trivial. The implementation might take some effort, but the conceptual idea does not require any research or trial and error. Embedding an application window in a browser window might not have been obvious back in 1994, but only because browsers weren't well understood. Normally, if you patent a technique or machine, it will almost always become at least somewhat obvious in retrospect, but here the issue is stronger. It's that the idea here was inevitable, once browsers grew to the point that they were used to view all sorts of things. It's as if someone tried to patent a handful of general browser strategies, without developing them, in the hopes that one of them would pan out and they could collect licensing fees. (Eolas did generate a working version of their embedding idea, but they didn't do anything to popularize it as part of the web.)

To try to put it another way, advances in software often lead to radical changes in how we use the software. Consider browsers, or wireless, for example. Radical change in use patterns often leads to demands for obvious new features, that previously were either impractical or simply irrelevant.

PTO review

Here's the PTO review upholding the Eolas '906 patent in 2005.

One of the things to get out of this document is the sense in which the PTO process favors the inventor; they are legally bound to interpret the inventor's claims in the most favorable light. This is because the assumption is (required to be) made that the PTO got the patent right in the first place; thus, the appeal is not neutral.

See Page 3 (p 7 of the pdf) for a discussion of "obviousness".

See also Page 11 (15) for a discussion of the Toye reference, in particular the discussion on Page 13 of an early X-windows version of Object Linking and Embedding. The NoteMail viewer could display active content: "any application that displays through an X-server can insert its output ... dynamically onto a notebook page through an embedded 'virtual window'". The patent re-examiner then asks whether NoteMail is "equivalent to" a browser.

Further on (Page 17/21) it is stated that

MediaMosaic does enable interactive control and manipulation of objects embedded in what arguably may be construed to be a "browser-controlled window," BUT ONLY AFTER USER INTERVENTION, such as by making a selection with a mouse. [capitals in original -- pld]

In other words, Eolas apparently has a patent on having something happen without having to click the mouse; their patent is for "zero-click" operation (one better than Amazon!). This was an important issue in 2003 as well. It is not a triviality. However, it is difficult to see how the invention of not having to click the mouse can be seen as deep. We're no longer talking about the idea of embedding applets; we're talking about automatically launching them.

Is not clicking the mouse worth half a billion dollars? That's more than the cost of every mouse in the US!

And finally, Eolas was derailed. In February 2012, Eolas lost a major District Court jury trial in Texas against Google, Yahoo, Amazon, JC Penney, CDW, Staples, and maybe one or two others. The jury held that all of Eolas' claims were invalid. Tim Berners-Lee testified, as did Pei-Yuan Wei (developer of Viola) and Dave Raggett (inventor of the <embed> tag). The defense took pains to argue that Doyle's idea was (a) obvious even in 1993 and (b) was not new; a 1991 Berners-Lee email (two years before the Eolas patent) was presented, in which he wrote "One thing I’d like to do soon, if I have time, is to teach the parser about Viola object descriptions and basically embed Viola objects into HTML files".

More at

Those who previously entered into agreements with Eolas to pay royalties must still do so.

In July 2012 the Texas court issued a ruling that the patents were invalid.

In July 2013, the Federal Circuit upheld the District Court ruling.

At this point the Eolas claims appear to be dead.

Why Texas? See


An earlier example of claims inflation is the E-Data patent, number 4,528,643. The original patent was granted in 1985 to Charles Freeny, for a kiosk-type vending machine for point-of-sale creation of audio cassettes or CDs. Such a device would be set up in music stores and would free the store from having to carry physical inventory. It is not clear, however, if any major music publishers agreed to allow their content to be sold that way; in any event, Freeny sold his patent in 1989 for $100,000.

E-Data bought the patent in 1994 and interpreted it as a device to sell merchandise using computers, thus raising the claim that it covered much of e-commerce. Specifically, E-Data tried to argue that the patent covered all online sales where the finished product was delivered via computer (examples might include e-books, digital music and other digital content).

Note that this is predated by the abstract idea of using computers to perform whatever business functions they can.

Multiple lawsuits were launched in 1995. In 1998, District Court judge Barbara Jones decided that the patent applied only to in-store sales, not online sales from off-site locations. This was a common-sense interpretation of the patent claims. The patent abstract is as follows (bullets added):

The present invention contemplates a system for reproducing information in material objects at a point of sale location wherein

The Federal Circuit then reversed this decision in 2000 (Interactive Gift Express v Compuserve, 99-1324), ruling that "point of sale locations" could include ones own living room, and that a home computer could be considered an "information manufacturing machine" as defined in the first formal claim of the patent (as the place where the reproduction occurred), and, for that matter, that a digital download could be considered a "material object". The Federal Circuit also ruled that the "authorization code" could be anything authorizing the sale, including an ordinary acknowledgment.

In 2003, Apple settled and entered into a licensing agreement. However, the terms of the agreement were not disclosed, meaning that Apple may in fact have paid a token fee (or perhaps even no fee at all); the existence of the settlement then gives ammunition for extracting settlements from other companies. The US patent expired in 2003, but judgments for past infringement were still a risk after that date.

Note that claim expansion is not an automatic win for the patent holder. Once E-Data got the claim-expansion ruling, defendants could introduce as Prior Art any evidence of selling ebooks or music downloads online, not just selling through in-store kiosks.

Two more specific patent cases

Patent cases we looked at:

Patent Trolls

Other names are Patent Holding Companies and Non-Practicing Entities, or NPEs. What to do about them? They do create a market for the sale of inventions. While many NPE patents come from bankruptcy sales and other sales, some still come from small inventors with no other way to capitalize on their patent.

Because NPE companies do not make anything, they cannot be countersued or convinced to enter into a reciprocal patent-sharing agreement.

If a company buys a "good" patent, like the RSA patent, licensing it is a socially reasonable business model. The problem, however, is with companies who purchase rather weak patents, and go after companies using the invention. This is, in part, a problem because of the issuance of a weak patent. But another part of the problem is that it is (or used to be) much cheaper to sue for an infringement claim than to defend against it, meaning that even if the patent was clearly inapplicable a company might still agree to settle.

The Obama administration issued legislative proposals and interim executive orders on June 4, 2013, intended to curb patent trolls. This is difficult because there is so much money on the table. Here is an abridged version of the executive order outline, from


Today the Administration is also announcing a number of steps it is taking to help bring about greater transparency to the patent system and level the playing field for innovators.  Those steps include:

  1. Making “Real Party-in-Interest” the New Default.  Patent trolls often set up shell companies to hide their activities and enable their abusive litigation and extraction of settlements.  This tactic prevents those facing litigation from knowing the full extent of the patents that their adversaries hold when negotiating settlements, or even knowing connections between multiple trolls....

  2. Tightening Functional Claiming.  The AIA made important improvements to the examination process and overall patent quality, but stakeholders remain concerned about patents with overly broad claims — particularly in the context of software.  The PTO will provide new targeted training to its examiners on scrutiny of functional claims and will, over the next six months develop strategies to improve claim clarity, such as by use of glossaries in patent specifications to assist examiners in the software field. [pld: about time!]

  3. Empowering Downstream Users.  Patent trolls are increasingly targeting Main Street retailers, consumers and other end-users of products containing patented technology — for instance, for using point-of-sale software or a particular business method.  End-users should not be subject to lawsuits for simply using a product as intended, and need an easier way to know their rights before entering into costly litigation or settlement. [pld: may not help much]

  4. Expanding Dedicated Outreach and Study.  Challenges to U.S. innovation using tools available in the patent space are particularly dynamic, and require both dedicated attention and meaningful data.  [pld: chatter]

  5. Strengthen Enforcement Process of Exclusion Orders. Once the U.S. International Trade Commission (ITC) finds a violation of Section 337 and issues an exclusion order barring the importation of infringing goods.... [pld: irrelevant to troll suits]

At least it is a start.

Some of the legislative proposals are more interesting:
  1. Require patentees and applicants to disclose the “Real Party-in-Interest,” by requiring that any party sending demand letters, filing an infringement suit or seeking PTO review of a patent to file updated ownership information, and enabling the PTO or district courts to impose sanctions for non-compliance.

  2. Permit more discretion in awarding fees to prevailing parties in patent cases, providing district courts with more discretion to award attorney’s fees under 35 USC 285 as a sanction for abusive court filings.

  3. Expand the PTO’s transitional program for covered business method patents to include a broader category of computer-enabled patents and permit a wider range of challengers to petition for review of issued patents before the Patent Trial and Appeals Board (PTAB).

  4. Protect off-the-shelf use by consumers and businesses by providing them with better legal protection against liability for a product being used off-the-shelf and solely for its intended use.

  5. Use demand letter transparency to help curb abusive suits, incentivizing public filing of demand letters in a way that makes them accessible and searchable to the public.

The best way to end abusive patent lawsuits, however, is to disallow such patents in the first place.

See also trolls2.

End-User Patent Lawsuits

A recent legal strategy has been for patent-holding companies to sue end users rather than manufacturers.

For one approach, see Austin Meyer's (warning: video autoplay with sound). Meyer is being sued for selling his app on the Google Play Store.

Apple lost a case to VirnetX based on specious patents, and now Apple network services may be in trouble:

Innovatio IP

Innovatio IP bought several Wi-Fi patents from Broadcom, and in 2011 began a large series of patent lawsuits against a large number of hotels and coffeeshops that provide Wi-Fi access for customers, and who are using routers from Cisco, Netgear and Motorola.

Innovatio was founded by Noel Whitley, a former Broadcom IP attorney. That more or less explains how the company was able to buy the patents.

The patents are not for the use of Wi-Fi by customers, or something silly like that. They appear to be genuine patents on Wi-Fi technology. The problem is that the manufacturers claim that the patents were legitimately licensed, but Innovatio is not suing them, and therefore they have no standing to defend the cases. The end-users who are being sued are not, generally, in a position to fight the lawsuits.

Cisco, Netgear and Motorola filed up to file a RICO lawsuit against Innovatio, but the judge did not allow that case to proceed.

When the IEEE set the Wi-Fi standard, all patent-holders who wanted their technology incorporated had to agree to license on RAND (Reasonable And Non-Discriminatory) bases (sometimes FRAND: Fair, RAND). But the RAND terms covered only what were considered to be "essential" patent claims (though this is a very broad category). Innovatio is arguing -- possibly speciously -- that the claims they are litigating are non-essential claims.

See more at

Project Paperless

Another set of companies sometimes collectively known as "Project Paperless" (but using a large number of shell-corporation names) has been sending out large numbers of letters to businesses claiming their use of office scanners to email pdf attachments is infringing. A sample letter is at

Some of the patent numbers are 6,185,590, 6,771,381, 7,477,410 and 7,986,426.

Quite aside from the question of going after end users is the question of licensing fees. The Innovatio and PP patents should probably involve fees on the order of $0.10, rather than $10,000.

Bessen and Meurer, Patent Failure

In 2008, James Bessen and Michael Meurer wrote a book Patent Failure: How Judges, Bureaucrats and Lawyers put Innovators at Risk, outlining their research on the hidden costs of the patent system. Somewhere in the last century Bessen was CEO of one of the first desktop-publishing companies.

One statistic that the book presents is the profit-to-litigation costs ratio, where both categories of costs are due to the existence of the patent. For pharmaceuticals, the ratio is typically 15:1. For other industries,  this ratio until the early 1990s hovered around 1:1, but in recent years has fallen to 1:4 (that is, litigation costs have outweighed profits fourfold!). For the specific industry of software, the issue is likely even worse. Be aware, though, that counting up profits associated with patents is difficult, and that arguably the goal of patents is to create incentive, not profits per se, and that unraveling the connection between incentive and profits is difficult. There are also many other social costs to patents besides litigation: suppressed invention is one.

Bessen and Meurer are particularly concerned with the intertwined issues of patent boundaries and notice. The patent boundary is the scope of the patent; that is, exactly what is covered. Patent notice is the act of making the boundaries apparent to others. Bessen and Meurer give examples (such as Eolas and E-Data) demonstrating that patent boundaries are highly subjective; they argue in fact that the lack of clear boundaries means nothing less than that patents are fundamentally not analogous to property: Without Boundaries, It Ain't Property is the title of their third chapter.

Furthermore, they argue that in most other examples of things considered to be property, boundaries and notice are crystal clear. Land is titled, and can be fenced and posted (with No Trespassing signs); in fact, if the land is not posted, you cannot prosecute someone for trespassing. Ownership of personal property is almost always very clear, also. Sometimes ownership of water rights is ambiguous, as are some forms of mineral rights. Patent boundaries are ambiguous because of:
  1. unpublished claims
  2. difficult-to-interpret claims (due to obfuscatory language)
  3. changes in claim interpretation (by the plaintiff or by the courts; E-data and i4i are good examples)
  4. very high search costs
Claim inflation is sometimes allowed because a judge feels an "impulse to award pioneer inventors". Even if that inventor didn't actually invent what they claimed!

Bessen and Meurer attribute the software-patent problem largely to the fact that software patents are so often patents on abstraction. Abstract claims are fundamentally broad; for example, the Supreme Court rejected in 1840 a patent for "the art of cutting ice by means of any power, other than human power". It is in the nature of an abstract claim to include things not actually invented. An abstract claim covers how things may be done in the future. Samuel FB Morse, in his patent for the electromagnetic telegraph, tried to make the following eighth claim:

I do not propose to limit myself to the specific machinery, or parts of machinery, described in the foregoing specifications and claims; the essence of my invention being the use of the motive power of the electric or galvanic current, which I call electro-magnetism, however developed, for making or printing intelligible characters, letters, or signs, at any distances, being a new application of that power, of which I claim to be the first inventor or discoverer.

This would cover the internet and the fax machine. Should it?

In the 1854 case O'Reilly v Morse, the Supreme Court threw out Morse's above eighth claim, and ruled that abstract ideas were not patentable. Alas, the Federal Circuit has tolerated some abstract claims.

One form of abstract claim is the functional claim: to claim any mechanism that achieves the same desired result.

The abstraction issue is not the same as the obvious-in-context issue. But the combination of abstract patents on obvious ideas (such as the E-Data patent on selling with the help of a computer) does have the effect of further expanding patent-holders' rights.

Bessen and Meurer suggest several possible reforms:
The first is widely recognized as an important strategy, and to a limited extent was incorporated into the Patent Reform Act of 2011. The second, however, would likely be much more important.

More recent proposals:
  1. Immediate publication of all patent applications (versus after 18 months, currently)
  2. Limit changes to claim language in pending patents (to prevent a pending patent from morphing into something else entirely). These changes, though, are actual changes to the claim text, made by amending the patent; this does not refer to changes in interpretation.
  3. Availability of greater information about prosecution history, that is, other infringement lawsuits filed for the same patent.
  4. Mandatory disclosure of changes in patent ownership
This last one may actually win approval, but it is very narrow.

Reading original patent descriptions

You can look them up online. Follow -> patents -> patft ( Or go directly to the search-by-patent-number page:

Google can also be used to search for patents at

State Street Bank case, and Business-Method patents

Patent 5193056

decided 1998; patent filed 1993
The patent was filed by Signature Financial Group on a mechanism for calculating taxes on certain pooled-fund investments. The basic outcome of the case was that financial-services software containing an algorithm became patentable. From the 1998 Federal Circuit decision:

(...) the transformation of data, representing discrete dollar amounts, by a machine through a series of mathematical calculations into a final share price, constitutes a practical application of a mathematical algorithm, formula, or calculation, because it produces 'a useful, concrete and tangible result' -- a final share price momentarily fixed for recording and reporting purposes and even accepted and relied upon by regulatory authorities and in subsequent trades.

For many decades, patents for business methods were generally denied (to be more precise, this "business-methods exception" to patentability was an oft-stated maxim, but the courts had never ruled directly on it). Even after the 1981 Diehr case allowing some software patents, a claimed software invention that also involved a business method might be refused a patent. By a decade after Diehr, however, the USPTO had begun granting a fair variety of software-related inventions of new business methods or business processes; it was too hard to make a distinction between software patents and business-method patents. When State Street Bank infringed on Signature Financial Group's patent, the latter sued, and SSB then tried to use the business-methods exception as a defense. In its ruling, the Federal Circuit established that there was no "business-methods exemption" as such, and that the USPTO had been within its rights to grant such patents.

The State Street Bank case introduced the rule that a business method could be patented if "it produces a useful, concrete, and tangible result". This "useful, concrete and tangible result" test was at the heart of the Bilski case, below, in 2009.

For a while, business-method patents were only granted for software-related processes; that is, business methods that involved the "technological arts". However, in 2005 the USPTO overturned this,  as there was no "technological arts" requirement written into law. The "useful, concrete and tangible result" rule was the only one that applied (note that this rule does pretty much forbid patenting a business idea). There remains a great deal of overlap between business-method patents and software patents.

Somewhere during this time the USPTO also reversed its longstanding refusal to grant patents purely for software (until then, software patents had to list some hardware involvement, although that was mostly a formality in practice). Whether this was related to the rise of business-method patents is unclear.

This case opened the door to a huge influx of business-method patents, including Amazon's "one-click" method (below).

Exhibit A: Amazon "one-click" patent, # 5960411, granted 1999. Twenty-three days later, Amazon got an injunction against Barnes & Noble, and eventually won their case. BN had to stop using their "ExpressLane" shopping method.

The essential feature of the 1-click patent appears to be that you don't have to go through the usual multi-screen checkout process; you can just click "send me this stuff the usual way", and it will remember all your previous entries.

Paul Barton-Davis, one of Amazon's founding programmers, called the 1-Click patent "a cynical and ungrateful use of an extremely obvious technology" [emphasis added]. By "ungrateful", Barton-Davis apparently meant that Amazon had benefitted enormously from public-domain software ideas.

One-click shopping is obvious from a software-development perspective. It is not nearly so obvious from the perspective of online shopping, however.

Overall, there are many other more-mundane patents on online shopping carts. IBM[?] supposedly has a patent on suggesting new purchases based on past ones.

See the Bilski case (later) for more updates.

Apple Patents

Just so you know who we're dealing with, here is a memo from Steve Jobs in 2010 in which he describes "Holy War with Google", presumably in large part because of Android. But he (later on in the memo, in item 5) says "Strategy: catch up to Android where we are behind (notifications, tethering, speech, ...) ...

Patent ruling in Apple "Cover Flow" feature. Apple purchased the patent from the developers in 2006, but another group (Mirror Worlds) also claimed a patent. But the judge ruled in Apple's favor.
album cover-flow image

Graph of phone-patent lawsuits

Apple seems intent on suing Android out of existence, targeting both Google (developer of Android) and all hardware phonemakers using Android, apparently believing Android to be a knockoff of the iPhone. However, history makes this much less clear. Smartphones were introduced in 2001 and 2002 by Palm, Microsoft and Blackberry. The iPhone did not appear until 2007. The iPhone did introduce the multi-touch interface, though by 2007 Microsoft Surface develpment was well under way. Android was introduced in 2008.

Similar lawsuits have been filed by Apple against makers of products competing with the iPad; the Samsung Galaxy tablet has been banned in Australia. But most of what Apple has are design patents, not software patents. Ironically, Apple already licenses several Samsung patents for phones generally. And many observers have commented that HTC seems to be introducing innovative features faster than Apple right now.

Apple filed a large batch of design patents in January 2007, starting a few days before the iPhone was introduced. Design patents for the Apple iPhone interface do make sense, in general; competitors should not be able to copy Apple's "look and feel". On the other hand, design patents are by their nature pretty straightforward to work around: all you have to do is make the competing product look different.

Apple based the Macintosh interface on work at Xerox PARC, and then failed to protect it from Microsoft (actually, the judge ruled that Apple had licensed the technology to MS). At least some of what Apple wants to protect in their iPhone cases is the physical design, though much of the design basics came from the work of Dieter Rams.

That said, most of Apple's innovations are genuine innovations, not involving elaborate claim expansion or subterfuge. Here, for example, is their "location" patent: 6122520 / RE42927: System and method for obtaining and using location specific information. Here is an article about it:

Apple has developed many very clever ideas for the iPhone. Many of these patents seem rather broad, however, especially the location patent.

Note that the strategy of Samsung/HTC was to file infringement counterclaims. This is why it is much harder for a legitimate developer to sue for infringement than a troll. However, the Samsung and HTC patents appear mostly to be general patents on wireless technology that they had previously agreed to license on Fair, Reasonable And Non-Discriminatory (FRAND) terms.

Multitouch patent:,2817,2387401,00.asp

Here's an amazing (though not up-to-date) graph by Florian Müller on Apple patent litigation relating to the iPhone:

Apple v Motorola

In June 2012, Richard Posner (normally of the Seventh Circuit but sitting temporarily as a District Court judge) dismissed an Apple v Motorola suit, arguing that Apple had largely failed to demonstrate the need for damages beyond a nominal level.

On November 5, 2012, Wisconsin District Court judge Barbara Crabb dismissed the Apple v Motorola case filed there, in which Apple accused Motorola of asking too much for its patents and was asking Judge Crabb to set a license amount. Motorola had asked for 2.25% of all net sales for use of its standard cell-phone patents. A problem is that while that may be a reasonable percentage for a "stupid" phone that doesn't do anything beyond place calls, it applies an unreasonable tax on "smart" phones with many non-cellular-related features. But the case dismissal apparently had more to do with the fact that the entire situation was more properly part of a larger case. Apple was asking Judge Crabb to act more as a mediator, and was not willing to be bound by any eventual decision.

Both of these cases are still going on; check in at Find a comfy chair first.

Apple v HTC

Apple v HTC before the International Trade Commission (ITC), filed March 2, 2010

5481721    Method for providing automatic and dynamic translation of object oriented programming language-based message passing into operation system message passing using proxy objects
5566337    Method and apparatus for distributing events in an operating system
5946647    System and method for performing an action on a structure in computer-generated data
5969705    Message protocol for controlling a user interface from an inactive application program
6275983    Object-oriented operating system
6343263    Real-time signal processing system for serially transmitted data

The ITC issued a partial ruling against HTC on December 19, 2011: HTC's phones could not use the "data-tapping" mechanism, in which phone numbers embedded in text (eg in email) are identified and the user can tap them to dial. This is described in the 5946647 patent, claims 1 and 8. Identifying embedded URLs by browsers and email clients has been around for a long time, although the mouse was needed to "tap" them.

While removing the data-tapping functionality might be a problem for HTC, Google may be able to develop a workaround.

The ITC ruling here suggests that, in the long run, Apple could prevail against HTC only on narrow claims, and the Android market is not in any real danger.

Apple v HTC, US District Court, Delaware

Apple also filed a conventional lawsuit against HTC in Delaware.

5455599    Object-oriented graphic system
5848105    GMSK signal processors for improved communications capacity and quality
5920726    System and method for managing power conditions within a digital camera device
6424354    Object-oriented event notification system with listener registration of both interests and methods
7362331    Time-based, non-constant translation of user interface objects between states
7479949    Touch screen device, method, and graphical user interface for determining commands by applying heuristics
7633076    Automated response to and sensing of user activity in portable devices

HTC counterclaims

5377354    Method and system for sorting and prioritizing electronic mail messages
6188578    Integrated circuit package with multiple heat dissipation paths
7278032    Circuit and operating method for integrated interface of PDA and wireless communication system

Apple and HTC settled all their lawsuits on November 10, 2012. At the moment, however, this is seen as an indication of HTC's relative weakness, compared to Samsung and Motorola.

In April 2012 Apple actually lost an important patent case in Germany, against Motorola, and so for the time being it's harder to buy an iPhone there, or use the iCloud service. But my understanding is that Apple is well-positioned to prevail in the long run.

Apple v VirnetX

In November 2012, Apple lost a patent infringement suit filed by VirnetX, for technology relating to FaceTime (and perhaps to some VPN-related features).

Apple v Samsung

This is Apple's big stand against Android. Samsung here is a stand-in for Google, which Apple was initially reluctant to sue because Google doesn't actually make any phones. (Apple may now be preparing a suit against Google.)

Samsung has also countersued, both in the US and throughout the world. Outside the US, Samsung met with a reasonable degree of success.

The first (of two) US Apple-v-Samsung lawsuit was filed in April 2011. The case involved both design patents and utility patents.

The utility patents were

Is the last patent a material improvement over having to zoom using the "pinch" gesture?

Apple also won on three design patents:

Design patent 504,889, on the rectangular case with rounded corners, was also included in Apple's complaint, but the jury did not find for Apple on this. It would be difficult to find a workaround for a rectangular case with rounded corners, as this layout is largely functional.

In August 2012, a jury in California returned a verdict against Samsung, with a judgment of $1.05 billion. Part of these damages were for design patents, including the infamous "rounded corners". But two key patents were the "edge-bounce" patent, 7,469,381, and the "tap to zoom" patent 7,864,163.

During the trial, Apple was able to present emails from Samsung indicating that they needed to make their phones more like Apple's and less like Nokia's. These were probably not internal calls for direct infringement, however, but simply awkwardly worded. Samsung at that point had simply realized that its main competitor was in fact Apple.

Each side was given a rather limited time to present its arguments.

Two months later, the USPTO issued a preliminary invalidation of the '381 bounce-back patent.

Later, questions were raised about some of the jurors. Jury foreman Velvin Hogan was a patent-holder himself. In an interview he claimed the jurors reached their damages award in order to punish Samsung, but the instructions to the jury were to set the award strictly for compensation for Apple's lost sales.

Both the edge-bounce and tap-to-zoom patents deal with very specific details. What about broad patents on the touchscreen interface? It seems unlikely at this point that Apple has those. What about patents on Apple's specific overall "look and feel"? It is not even clear Apple has those, although if they did workarounds should be straightforward. The edge-bounce, tap-to-zoom and slide-to-unlock (below) patents are realistically all really of this kind: patents on how the user interface works. If they are interpreted narrowly, there is no problem. If they are interpreted broadly (eg interpreting the slide-to-unlock patent as forbidding any sliding motion related to unlocking), there is reason for concern.

Eventually a retrial (at least of the damages) was ordered, and the damages were reduced in November 2013 to $290 million.

In May 2015 the Federal Circuit issued its appeal ruling, upholding the design-patent infringement claims (and also some non-patented trademark claims). Samsung had argued that (a) mostly they weren't violating Apple's design patents (a Samsung S3 cannot be mistaken for an iPhone, after all), and that (b) to the extent that they did copy Apple's design, it was for functional purposes only (you can't patent function). The Federal Circuit didn't buy either argument.

Worse, for Samsung, the 1952 law covering design patents (35 U.S.C. § 289) states that all profits from the infringing design be surrendered: "Whoever during the term of a patent for a design ... applies the patented design... shall be liable to the owner to the extent of his total profit". There is no "proportionality" rule covering cases where most of a device (eg a phone) is not infringing. The law was presumably written with situations in mind in which the applicable design was the primary selling point (eg the 1950s-era Eames chair).

This rule might make sense if the design infringement was narrowly construed, eg if the infringing design had to intentionally mimic the original. It seems unlikely Samsung was doing that; all phones look superficially alike.

In September 2015 the Federal Circuit issued a revised damage award of $548 million (up from the 2013 $290 million). In December 2015 Samsung agreed to pay this, but reserved the right to ask for the money back if a later appeal succeeded; see

On March 21, 2016, the Supreme Court accepted Samsung's appeal ( The Federal Circuit's interpretation of design patents, and the 1952 law above, required forfeiture of all profits for infringement. In December 2016 the Supreme Court invalidated that full-forfeiture rule, and sent the case back to the Federal Circuit to determine how much money should be owed for Samsung's infringement on Apple's design patent.

Apple filed a second lawsuit against Samsung in February 2012. Jury selection began on March 31, 2014. The utility patents at stake are
This trial includes the still-more-or-less-current Galaxy SIII (though the S7 is now out).

Apple worked hard on integrating all the features above into a consistent iPhone interface. But some of the patent claims are decidedly unconvincing, especially '414.

In May 2014 Samsung was ordered to pay Apple $119.6 million in this case, an amount widely regarded as much smaller than Apple was hoping for.

In February 2016 the Federal Circuit overturned even this modest award ( Apparently the Federal Circuit actually threw out the relevant Apple patents, calling them "obvious", but Apple can still appeal.


One famous Apple patent, not used in the Apple v Samsung case above, is : slide-to-unlock. In its original form (8,046,721), the slider had to appear at a specific point on the screen, and be moved "to a predefined location and/or ... along a predefined path". Apple has now received a couple more slide-to-unlock patents; the latest is 8,286,103. This latest patent does still refer to "unlock images":

The device is unlocked if contact with the display corresponds to a predefined gesture for unlocking the device. The device displays one or more unlock images with respect to which the predefined gesture is to be performed in order to unlock the device.

The ice-cream-sandwich (Android version) slide-to-unlock mechanism involves no unlock image whatsoever; one simply drags a finger along the screen. The action creates ripples; if the path is long enough, the phone unlocks.

Is this a legitimate workaround?

How is slide-to-unlock different from the very standard GUI action of dragging an object?

Slide-to-unlock is a pretty central patent for smartphones; life would be much harder if you had to enter an unlock code each time. Slide-to-unlock is also a feature with no physical-world analog, unlike many other GUI features.

One of Apple's slide-to-unlock patents was invalidated in Germany in April 2013, largely (if not entirely) because software is not patentable in Europe unless it "solves a technical problem with technical means" [].

Very recently, there is new evidence that the slide-to-unlock gesture was developed ten years before Apple patented it:

Florian Müller's excellent patent blog has a good summary of what he considers Apple's top four patents at

Müller makes an argument that the first one is the most important, because it is revolutionary and not evolutionary. Past practice had been to display a document with "hard" margins. Müller writes:

Rubber-banding solves the problem that an abrupt end to a scrolling operation instinctively makes a user press harder because it appears that a device is not responding to a gesture. ... But from an end user's perspective, rubber-banding is a solution to a problem. Having a "glow" at one edge of the screen is better than getting no visual feedback at all, but it's too subtle.

(Personally, I think edge-glow is just fine, but I get his point.)

The '949 patent is interesting:

detecting one or more finger contacts with the touch screen display, applying one or more heuristics to the one or more finger contacts to determine a command for the device, and processing the command.

It is at the heart of what makes the iPhone touch interface so effective. And yet the courts have so far not been impressed with it. Apple obtained an injunction banning the import of devices violating it, but Android has found pretty effective workarounds; see The USPTO has also issued a preliminary invalidation.

The 7,844,915 version of the pinch-to-zoom patent has received preliminary invalidation in July 2013.

Google and Microsoft sue GeoTag
GeoTag appears to be a classic troll. But the article mentions Skyhook, and suggests that is not a troll. Does this matter? What about i4i?

Here's the GeoTag patent: 5,930,474

How this compares with RE42927 I have no idea.

This patent is definitely an example of an idea, the implementation of which is obvious, given the availability of GPS. Should such things be patentable?

Stallman article: why software is different

1. There's no advance warning (but isn't this true of any patent?)

2. There is no easy way to read them. They are deliberately obscure. And, for software, this is directly tied to the fact that the algorithms are very general.

3. Some patents are just plain inappropriate, but fighting them is exhorbitant.

4. chicken-and-egg problem with converting from .gif to .png: how can you get browsers to support this? (This is partially resolved by browser plug-ins.)

5. "prior art" is very hit-or-miss; sometimes (often!) it was considered too obvious to document. More specifically, ideas ruled by the PTO to be patentable were often "CS folklore" but too obvious to publish.

6. Software tends to use many ideas per application, and so one software project might require licensing of many diffferent pieces. This makes incremental innovation difficult, for anyone. (Even microsoft.) This problem exists for software because writing software is in some sense much easier than traditional engineering, and so you can afford to put more ideas in.

7. Software licensing terms tend not to take into account the fact that projects use many ideas, and a given patented piece should be only a small part of the total.

8. Open source.

9. Patented standards are a large social cost.

10. It is not possible to search the patent database before the patent is granted

11. Tremendous software progress is often made with negligible investment; we don't need patents to encourage development.

12. Computing has a high rate of radical, as opposed to incremental, change.

An issue that Stallman doesn't address directly is that 20 years is a very long time in the software industry, and that locking an idea up for that long has the practical effect of stifling growth, not encouraging it. Software patents have had the effect of postponing adoption until patent expires (cf James Gosling's first language) (hint: his second language was Java).

And basic ideas are still being patented for a new use (eg the Steir hair patent)

Finally, another issue is the obvious-in-context problem I mentioned above, that does seem to be a special issue with software: often an idea is an immediate, trivial corollary of some later development, or of some more abstract way of looking at the situation. However, some earlier out-of-context patent may still apply. The XOR-cursor and Eolas patents may be of this type. It's not that embedded applets is an innovative idea, per se, it just takes a critical mass of web sites to become relevant.

Is software legitimately a special case?

A related issue is that weak patents can have significant social costs, by blocking core technologies that everyone needs. Examples are not always easy to identify.

It is more likely for workarounds to be found, or for a feature to be licensed and added to non-open-source software.


Standards organizations often require that patent holders who want their technology incorporated into a standard agree to license the technology on "fair, reasonable and non-discriminatory" terms (this is also sometimes abbreviated RAND).

The theory behind F/RAND licensing is that the patent should be reasonably inexpensive, and no company should get a special deal. Essentially all the core cellphone technology is licensed under F/RAND terms.

Apple v Qualcomm

F/RAND makes patented standards a little more reasonable, but does not resolve all the problems. In 2017 Apple sued Qualcomm, owner of a large number of cellular and mobile-internet patents, over its licensing terms. Qualcomm's interpretation of "fair and reasonable" is that it charges a percentage of the phone's price, around 5%. For a $400 phone, that's $20. But for a high-end $1000 iPhone, it's $50.

Qualcomm allows chip makers to manufacture chips based on Qualcomm's technology for free, more or less. But the company that buys the chips and builds a consumer product has to pay the 5% licensing "tax".

Apple's issue is that most of the costs of an iPhone are not related to the communications technology. Apple has a large number of innovations, and feels that Qualcomm's licensing fee should be a flat rate. Qualcomm is profiting from every non-telecom feature Apple adds to the iPhone.

Apple has a point. Apple's case is bolstered by a Supreme Court decision disallowing royalties if a part is sold outright; the patent-licensing costs must be entirely included in the selling price of the part.


For another tale of F/RAND woe, see Alastair Houghton developed in 2014 a way of connecting speakers to Apple laptops so as to get digital surround sound. There was one catch: the software made use of the Dolby AC-3 codec, which was patented.

Dolby licensed the technology under F/RAND terms, though, and initially Houghton thought he could work out a deal. But the set patent terms assumed a mass-market level of sales, and were much too high for Houghton's niche product.

Dolby considered Houghton's request for special licensing terms, but in the end decided they could not do that because that would violate the non-discriminatory part of F/RAND. Not everyone would get the same deal! The pricing structure built into the F/RAND terms made sense only for hardware devices, not for software.

The patent for AC-3 finally expired, and Houghton was able to release his package in 2018, four years later.

In 2008, the FTC weighed in to enforce the original F/RAND terms for licensing Ethernet speed auto-negotiation from N-Data, which had purchased the patents and tried to raise the licensing terms. See

Paul Graham

This would be a good time to take a quick look at the Paul Graham paper. Graham is both a venture capitalist and a software engineer (and a Lisp programmer!). One of his first points is the following:

One thing I do feel pretty certain of is that if you're against software patents, you're against patents in general. Gradually our machines consist more and more of software. Things that used to be done with levers and cams and gears are now done with loops and trees and closures. There's nothing special about physical embodiments of control systems that should make them patentable, and the software equivalent not.

The contrary position here is that physical systems are different. If you are building a complex machine with gears, you have to worry about all sorts of things: the strength of the gears, the precise shape of the gear teeth, whether the gears are meeting at an angle (as in bevel gears), whether there are pinion gears, the force between the gears, lubrication, temperature, and bearings for the gear shafts. None of that matters for digital systems. Gears, ultimately, are used to transfer power and to adjust "impedance"; those issues don't even exist for digital systems.

Suppose one is designing a mechanical valve. There are again all sorts of mechanical details to be addressed. Now consider a network router as a kind of "packet valve". To open the valve, this is all you need:

    ipforward = True

To close it:

    ipforward = False

(The "ipforward" name is used for this purpose in Linux; it is in /proc/sys/net/ipv4.)

The bottom line is that, yes, there are very advanced digital analogues of physical machines that are not trivial, but the most common problem with implementing software control systems is simply the development time. There is no "trial and error" to speak of.

Here are a few more Graham points, based largely on his experience as a venture capitalist. Patents, he feels, don't matter much to software startups, unlike physical-machine startups.

Graham has three reasons why patents don't matter:
  1. Software is complicated; the real issue isn't the software but developing -- and designing -- it. However, this argument also works the other way; if you have an idea, then you are better off pursuing patent enforcement than development, because development is hard. But also note Graham's point that if a big company tries to copy a little company's patents, there will be a "thousand little things the big company will get wrong".
  2. Startups seldom compete head-to-head with big companies; they "change the paradigm". You don't go into the word-processing business; you invent Writely (now part of Google Apps?). And, "big companies are extremely good at denial". They will go to great lengths to pretend that you don't exist, to "keep you in their blind spot". Suing a startup would mean you realized they were dangerous. He cites IBM as an example; it would have been demeaning for them to sue microcomputer developers. Also, for Microsoft to sue web-app developers (or smartphone developers) would be to admit that Windows is fading.
  3. Hacker opinion is against big patent lawsuits. If you're a big high-tech company, you'll lose a lot of your best people if you're seen "doing evil". This might be true for Google; it's less clear at Microsoft (though the employees there do care about principles). It's probably not true at Eolas.
What do you think of these? How does the Eolas case fit in?

The argument is sometimes made that patents have real value for startups, so that they can cross-license if accused of violating other patents. Is this a legitimate argument in favor of patents?

Graham is perhaps close to the heart of the matter on an other issue: computing startups do not usually get sued until they are large enough to attract attention. Does this mean they don't need patents?

There are anecdotal reports that some startups feel someone else would take over their idea if they did not have patent protection. How important is such protection, versus the "troll tax"?

Note that patents are for the use of an idea in a specific context:

Alas for computing, the PTO has often accepted patents for the use of a very standard, well-known idea in a specific context.

Patent problems

submarine patents: you don't hear about them until too late!

prior art: hard to find, hard to document, trivial ideas were never written down! This problem, at least, will go away with the passage of time.

non-obviousness: difficult to contest.

Many ideas go into one program! Technology evolves extremely rapidly.

Violates settled expectations (important part of law!) What's patented seems to be more a matter of chance than anything else.

ignorance is no defense: "submarine" patents entire process is secret: you can be making good-faith effort to be noninfringing and get hit with a huge verdict.

willful: you had advance notice of infringing. Your belief that the patent was invalid may NOT be a defense, although it has been accepted as a defense in some cases. Damages automatically triple.


EU Parliament voted in July 2005  648-14 AGAINST the EPO (European Patent Office) directive.

March 17, 2009: European Patent Office has asked the EU's "Enlarged Board of Appeal" to decide on the exclusion of software from patentability. The EPO has long been pushing for software patentability, and this is seen by some as an attempt to bypass the European Parliament.
Also Note especially Q3, under Questions. Under some earlier rulings (T163/85 and T190/94), patentability required "a technical effect on a physical entity in the real world". However, other rules did not include this requirement.

European patent law is similar to the Diamond v Diehr standard: machines that use software are patentable, but not software that stands alone. However, in the US the Diehr standard evolved into software patentability; in Europe software remains unpatentable as such.

Here's an article from entitled, "Why are Software Patents so Trivial?", in which they suggest that this is a fundamental problem:

In February 2016 the Indian Patent Office adopted a similar rule, that software alone is not patentable. See


Who are the stakeholders in software patents? Are we stakeholders? Compare pharmaceuticals. "Do you feel helped by patent reform?"

If the Eolas patent had succeeded earlier in the game, Firefox might never have been started, and then Internet Explorer would still likely lack tabs, plug-ins, and other core features.

WHY does the situation seem so different from pharmaceuticals?

Role of "patent trolls", or patent licensing firms
("troll" as in "the troll under the bridge, demanding tolls", not "trolling" as in fishing for "flames")
Note that the established-company-versus-established-company defense of a "patent bank" is useless against trolls that do not actually make anything.

Microsoft is both patent-troll victim (see the Eolas case, above) and patent troll itself. See, for example,

There is, however, one thing that many open-source people agree keeps them from really trusting Microsoft: Its continued demand that Android vendors pay them for Android patents.
...most of Microsoft's Android patents cover ideas that are "already part of the public domain."
This latter role is perhaps why Microsoft and other large companies has resisted patent reform.

Barriers to entry

Patent Trolls: companies that have no assets but patent claims, and don't attempt to produce anything but simply collect. Is this bad? Or are such companies just creating a market for small inventors to sell their inventions?

I4i is not such a company; they did produce an XML-based product.

Patent and open source

The open-source community is a strong proponent of eliminating software patents.

Is the open-source community entitled to:

Is the open-source community entitled to the Asterisk phone switch?

Does MS intend to destroy or hobble or marginalize linux through patents?

It is very well documented that the patent process can have a very NEGATIVE impact on open-source development, and on generally accepted software adoption.

So if the purpose of software patents is to aid technological process, and it doesn't do that, are software patents a good idea?

What happens if the software in question is made available through a site in Europe, which (as of now) doesn't have strong software-patent laws? Should the site warn visitors from the US?

Is this at all like

Patents: are the right ideas being patented? Or are patents being granted to trolls for peripheral ideas?

xor:   trolls?
rsa:   good
spreadsheets:   trolls?
eolas:   trolls?

DataTreasury (

They have developed technology for storage of digital images of bank checks. They actually did develop the whole system, although again the inevitability issue arises here. They did not develop any of the actual root technology: scanners, or data security, or digital storage systems with enough capacity to hold images for negligible cost.

From their website:

The Corporation was founded in 1998 and was granted its first two Network architectural patents (5,910,988 and 6,032,137) in 1999 and 2000, respectively. The patents detail the important and revolutionary aspects of DataTreasury's systems for remote image capture, document imaging, centralized processing and electronic storage. Our innovations were particularly noted for enhanced security, fault tolerance and high reliability. These key elements form the underpinnings of DataTreasury's technology.

That said, it is clear that none of DataTreasury's ideas are revolutionary.

From The company had benefited from a controversial 1998 court ruling that broadened the definition of a patent to include business processes.

The proposed (but never passed) patent-reform act of 2007 singled out this patent for congressional revocation.

It appears that DataTreasury is claiming a business-method patent on the use of electronic image scanning for check processing. They are looking for very significant licensing fees. Again, every piece of the technology has been around from well before the patent (scanning, secure storage, ???)

Should a new (but straightforward) application of existing technology be patentable?

The DataTreasury patent has been singled out by Congress for action, but it is not clear what will happen.

Patent reform:

Someone tried patenting a movie storyline a few years ago. This patent WAS rejected.

Patent Reform Act of 2007: H.R. 1908 and S. 1145 (this did not pass). This was the first [?] patent-reform act proposed.

Those in bold are the most significant.

This did not pass.  Here are some of the proposed changes in U.S. patent law

Discuss: first-to-file: who benefits? how are small inventors affected? How are prior-art rules affected?

This has again been introduced in 2009; apparently the issues are the damages calculation, post-issuance reexamination proceedings, and defining inequitable conduct. At least the last provision has been removed from the 2009 bill. A good-faith defense for believing a patent was invalid is also included. Also included is a definition of prior art to include anything "available to the public"; publication no longer would have to occur.

[Note that NTP argued that RIM's conduct was held to be inequitable simply because NTP had sent them a letter outlining its patent claims, and RIM had disagreed.]

In 2011, Congress passed the America Invents Act. This included the following features:

The full impact of the law is not yet clear. There were no changes to rules for willful infringement, inequitable conduct, access to injunctions, or proportionate damages.

KSR v Teleflex, Supreme Court, April 30, 2007

This Supreme Court case altered the legal standard for disproving "non-obviousness" in favor of defendants. It is now quite a bit easier to challenge patents on this basis. In particular, KSR v Teleflex makes it much easier to argue that software patents are simply a combination of pre-existing ideas; all the defendant has to come up with are the prior art for the pre-existing inventions and some rationale for why it would have been obvious to combine them. The defendant does not have to prove that it would have been obvious which pre-existing inventions to combine.

Teleflex had a patent on a pedal coupled to an electronic throttle control (basically cruise control). The question was whether that was "obvious".

The proper question to have asked was whether a pedal designer of ordinary skill, facing the wide range of needs created by developments in the field of endeavor, would have seen a benefit to upgrading [a prior art patent] with a [throttle-control] sensor

not thought of it by themselves, and not motivated to implement the change, but simply saw the benefit. Furthermore, the conclusion of obviousness needed only to be supported by ordinary evidence, not written documentation. The old "nonobviousness" standard often in effect required proving that a patent was "prior art". This test was known as the "teaching-suggestion-motivation" test. All three pieces had to be there. Another sentence from the KSR decision:

[t]he combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results.

By comparison, my obvious-in-context approach means that the idea (or combination) is obvious to anyone facing a later  Does that suggest that not clicking the mouse is obvious?

The KSR decision has made it straightforward to establish the obviousness of an invention that consists of two parts if

Would this have helped RIM? Perhaps. The Supreme Court did have an older obviousness test from the Graham case of 1966. In it they advocated considering the general scope of prior art, the extent to which the invention departed from prior art, the prevailing degree of ordinary skill in the field (eg, if software engineers were generally able to fit data structures to a given problem then a new application of a data structure might be "obvious"), and what the court called "objective" evidence. Unfortunately, the objective-evidence examples for non-obviousness included such things as "commercial success" and "long-felt but unserved needs": if the new product had great commercial success, and satisfied long-felt but unserved customer needs, it might be presumed non-obvious. That is exactly the test the RIM court applied to rule in NTP's favor. Alas, there are good reasons to suppose that RIM's commercial success had nothing to do with innovation, and everything to do with lining up investment in what was, at the time, a rather speculative need. (The third SCOTUS example of non-obviousness was the "failure of others" to solve the problem at hand; if that test were applied to most software patents, few would survive.)

The KSR case is apparently one of the most-cited Supreme Court patent cases.

Bilski case

Federal Circuit decision released October 30, 2008
Supreme Court decision released June 28, 2010 (decision here)

This was a very significant case. It was decided at the appellate level by an en banc sitting of the Federal Circuit. They proposed a "machine or transformation" test for patentability of abstract processes. The Supreme Court then heard the case, and while they did not uphold the "machine or transformation" test, they ruled that Bilski's invention was not patentable because it was too abstract. There had been widespread speculation that the Supreme Court would use the Bilski case to rein in business-method patents, or at least make the patentability rules a little clearer. They apparently did not do either.

Bilski patent: Claimed method of managing the risk of bad weather in commodities trading.

He submitted a patent application seeking exclusive rights to a method of using hedge contracts to reduce the risk that a commodity's wholesale price might change.

Again, the technique fails under both prior-art and obviousness standards. But those don't apply in the same way to business-method patents.

The patent was rejected by the Patent Board of Appeals. The Board, in rejecting the claim, asked the Federal Circuit court for assistance in determining patentability of non-technological method claims.

The federal circuit court did the following:

The court by its own action grants a hearing en banc. The parties are requested to file supplemental briefs that should address the following questions:

  1. (1) Whether claim 1 of the 08/833,892 patent application claims patent-eligible subject matter under 35 U.S.C. §101? (the patent-eligibility rules)
  2. (2) What standard should govern in determining whether a process is patent-eligible subject matter under section 101?
  3. (3) Whether the claimed subject matter is not patent-eligible because it constitutes an abstract idea or mental process; when does a claim that contains both mental and physical steps create patent-eligible subject matter?
  4. (4) Whether a method or process must result in a physical transformation of an article or be tied to a machine to be patent-eligible subject matter under section 101?
  5. (5) Whether it is appropriate to reconsider State Street Bank & Trust Co. v. Signature Financial Group, Inc., 149 F.3d 1368 (Fed. Cir. 1998), and AT&T Corp. v. Excel Communications, Inc., 172 F.3d 1352 (Fed. Cir. 1999), in this case and, if so, whether those cases should be overruled in any respect?

The appellate court did affirm the need for a physical transformation. Their central doctrine is "Machine or Transformation". This would have been a problem for business patents, and perhaps software patents.

Note that their reasoning was taken straight from the few SCOTUS cases on record.

The following question arises whenever a patent is applied for on an abstract process: 

[Is the patent] tailored narrowly enough to encompass only a particular application of a fundamental principle rather than to pre-empt the principle itself?

Benson: NO
Diehr: YES (one of the prior SCOTUS cases)
Bilski: NO

This part of the Federal Circuit's reasoning may still stand.

Part of the Benson ruling:

Transformation and reduction of an article 'to a different state or thing' is THE clue to the patentability of a process claim that does not include particular machines.

The Diehr patent was for making rubber, using a computer to control the process. It wins the "different state or thing" standard hands down.

The federal circuit dismissed the "useful, concrete, or tangible result" test: that is NOT enough to establish patentability.

They also reject the "technological arts" test (see above) that was once-upon-a-time part of the method-patent rules. They agree that it is too hard to tell whether something involves the technological arts; however, unlike the USPTO, they end up ruling the OTHER WAY; that is, to reject MORE broadly than the TA test.

machine-or-transformation test: emphasize the OR.

We will, however, consider some of our past cases to gain insight into the transformation part of the test. A claimed process is patent-eligible if it transforms an article into a different state or thing. This transformation must be central to the purpose of the claimed process. But the main aspect of the transformation test that requires clarification here is what sorts of things constitute "articles" such that their transformation is sufficient to impart patent-eligibility under §101.

Tanning leather curing rubber (Diehr case)

The raw materials of many information-age processes, however, are electronic signals and electronically-manipulated data. And some so-called business methods, such as that claimed in the present case, involve the manipulation of even more abstract constructs such as legal obligations, organizational relationships, and business risks. Which, if any, of these processes qualify as a transformation or reduction of an article into a different state or thing constituting patent-eligible subject matter?

Note that while the Bilski decision does not claim to reverse State Street (the case that led to business-method patents), most commentators seem to feel that it has that effect. It is less clear that Bilski would have had a significant effect on software patents.

Applying the Machine-or-Tranformation test to famous cases

RSA? material transformation in "real" terms The transformation is to a file. While it is electronic, it is decidedly material.

MP3? material transformation in "real" terms? An mp3 file isn't a physical thing, but it does have a certain "thingness". People think of them as things, and buy them as things. An mp3 file is material.

NTP? maybe no? The argument can be made that there is no "material thing" on the table here. Email messages are NOT it; the patent only addresses the delivery of email.

DataTreasury? It seems unlikely that DataTreasury's patents would stand up to this new test.

Supreme Court

Pamela Samuelson, writing in the March 2010 CACM, noted that the Supreme Court appeared during oral arguments to believe that some way was needed to disallow patenting of nontechnological processes. Justice Scalia asked whether horse-training techniques should be patentable, and techniques to "win friends and influence people". Justice Sotomayor asked whether speed-dating methods could be patentable, and Justice Breyer asked if a professor could patent an improved teaching method.

However, this did not quite happen. Here are a few quotes from the decision, written by Justice Kennedy [emphasis by pld]:

Section 101 specifies four independent categories of inventions or discoveries that are patent eligible: “process[es],” “machin[es],” “manufactur[es],” and “composition[s] of matter.” “In choosing such expansive terms, . . . Congress plainly contemplated that the patent laws would be given wide scope”

This Court’s precedents provide three specific exceptions to §101’s broad principles: “laws of nature, physical phenomena, and abstract ideas.”

The machine-or-transformation test is not the sole test for patent eligibility under §101. The Court’s precedents establish that although that test may be a useful and important clue or investigative tool, it is not the sole test for deciding whether an invention is a patent-eligible “process” under §101. In holding to the contrary, the Federal Circuit violated two principles of statutory interpretation: Courts “ ‘should not read into the patent laws limitations and conditions which the legislature has not expressed,’ ” Diamond v. Diehr, 450 U. S. 175, 182, and, “[u]nless otherwise defined, ‘words will be interpreted as taking their ordinary, contemporary, common meaning,’ ”

Section 101 similarly precludes a reading of the term “process” that would categorically exclude business methods. The term “method” within §100(b)’s “process” definition, at least as a textual matter and before other consulting other Patent Act limitations and this Court’s precedents, may include at least some methods of doing business.

Because petitioners’ patent application can be rejected under the Court’s precedents on the unpatentability of abstract ideas, the Court need not define further what constitutes a patentable “process,” beyond pointing to the definition of that term provided in §100(b) and looking to the guideposts in Benson, Flook, and Diehr. Nothing in today’s opinion should be read as endorsing the Federal Circuit’s past interpretations of §101. [that is, the Supreme Court is not endorsing the State Street Bank case -- pld]

The appeals court may have thought it needed to make the machine-or-transformation test exclusive precisely because its case law had not adequately identified less extreme means of restricting business method patents. In disapproving an exclusive machine-or-transformation test, this Court by no means desires to preclude the Federal Circuit’s development of other limiting criteria that further the Patent Act’s purposes and are not inconsistent with its text.

In other words,

We can kick this patent out without resorting to a Machine or Transformation test, and while we probably think the MoT test is too restrictive, we're not going to say anything further.

Return to Gottschalk v Benson (which Bilski v Kappos did apparently firmly uphold)

It is easy to interpret Bilski as reinforcing the Benson decision. It is up to the Supreme Court, however, to decide if Benson was in fact the right approach. The idea expressed in Benson that the algorithm was "too general" and might be used for anything seems in hindsight rather quaint; it is clear a few decades later that this is going to be the case with perhaps the majority of software patents. For example RSA patented a method of encryption that could be used for anything: banking, personal matters, commerce, digital signatures, etc.

Also, Section VI of a separate opinion by Stevens, Ginsburg, Breyer and Sotomayor is a strong argument against business-method patents, period.

The constitutionally mandated purpose and function of the patent laws bolster the conclusion that methods of doing business are not “processes” under §101 [of the Patent Law].

However, we will have to wait for a future case to see if this position makes it into an actual decision.

Mayo Labs v Prometheus Labs, Supreme Court, March 20, 2012

Prometheus Labs patented a way of measuring levels of thiopurine drugs to maintain a consistent dosage. The set of metabolites is measured, and from these varying levels an inference can be made as to the level of the original drug.

Mayo Labs used the patented idea without a license. Prometheus sued. The District Court found in favor of Mayo (more or less), but the Federal Circuit held that Prometheus's strategy was patentable, and, furthermore, even met the stronger "machine or transformation" test of patentability. The Federal Circuit's argument was that administration of the drug "transformed" the body, and that administering the blood test "transformed" the blood sample (my first reaction here is that it seems that similar logic could make anything fit the "machine or transformation" logic, thus undermining the whole purpose of the MoT test as a limitation on patentability).  Following Bilski back in 2008, the Supreme Court remanded the Mayo v Prometheus case back to the Federal Circuit for reconsideration; while the Bilski decision was often seen as expanding the rule of patentability beyond the MoT test, what the Supreme Court apparently had in mind was that sometimes things could be patentable while not meeting the MoT test but other times perhaps things could meet the MoT test and yet not be patentable. The Federal Circuit ruled again that Prometheus's  invention was patentable under the MoT test.

The Supreme Court ruled in March 2012

Because the laws of nature recited by Prometheus’ patent claims—the relationships between concentrations of certain metabolites in the blood and the likelihood that a thiopurine drug dosage will prove ineffective or cause harm—are not themselves patentable, the claimed processes are not patentable unless they have additional features that provide practical assurance that the processes are genuine applications of those laws rather than drafting efforts designed to monopolize the correlations.

What other patents that we've looked at might be construed as patenting general scientific principles? The eolas patent? Any patent where as standard data structure is applied to a specific context?

The Prometheus patent was denied on the grounds that there was nothing substantial to the invention beyond a fundamental scientific principle (relating to a blood test for the effective levels of a drug). The patent is for the idea that by measuring two metabolites of the drug the effective level can be determined.

For software patents, in many cases the anti-patent argument is that a claimed invention is nothing more than a fundamental principle, applied in a specific context.

Myriad Genetics

The Supreme Court also in 2012 sent back to the Federal Circuit court for reconsideration, in light of the above case, the Association for Molecular Pathology v. Myriad Genetics case. Here, Myriad Genetics had been issued a patent on two human genes, BRCA1 and BRCA2. Patenting of human genes is controversial because, well, perhaps if you had one of those genes you would have to pay a royalty? (The answer is no, of course; the patent would be for the use of the gene in a particular in vitro manner.)

The Federal Circuit upheld the patent in August 2012, and the Supreme Court heard the case. They ruled on June 13, 2013 that human genes can not be patented; that isolating a human gene is an act of scientific discovery rather than invention. This invalidated a large number of existing gene patents.

It does not affect patents for genetically modified organisms, though.

Myriad had not allowed others to test for the BRCA1 and BRCA2 genes, even with more sophisticated tests than their own.

Myriad's patent on synthetically created DNA still stands.

Abstract Patents

This is an important distinction that merits greater emphasis. In theory, "abstract ideas" are not patentable. In practice, the Federal Circuit has undermined this with a number of decisions, notably the State Street Bank case.

There is some uncertainty as to what, exactly, constitutes an "abstract patent", and when an idea is too general to patent. Bilski's process was one such abstract idea, but note that in the State Street Bank case the Federal Circuit had agreed that similar abstract ideas were patentable (to its credit, the Federal Circuit reversed course on that in their original Bilski decision).

In Ultramercial v Hulu the Federal Circuit has allowed patentability of the idea of requiring a subscriber to view online ads before viewing copyrighted content. The patent in question is 7346545. The abstract states

The present invention is directed to a method and system for distributing or obtaining products covered by intellectual property over a telecommunications network whereby a consumer may, rather paying for the products, choose to receive such products after viewing and/or interacting with an interposed sponsor's or advertiser's message, wherein the interposed sponsor or advertiser may pay the owner or assignee of the underlying intellectual property associated with the product through an intermediary such as a facilitator.

There are no techniques described in the patent. That is, the patent is on the abstract idea of showing ads on the internet.

This patent is also troubling on prior-art and obviousness grounds; advertising has been present on commercial broadcast TV from the beginning.

Such patents are also sometimes called functional patents: the patent describes only the "function" of the invention (showing ads, sending characters, cutting ice) rather than a specific method. Functional patents can also be considered a more specific category; some patents considered "abstract" do at least attempt to outline a method.

There is a theory that the most egregious problems with software patents would be fixed if abstract/functional patents were disallowed. Of the patents we have considered, the XOR patent, the spreadsheet-recalculation patent, the MP3 patents and the RSA patent are not functional/abstract; each involves a specific way of doing things. The Eolas patent may be considered to be functional, in that it covers any mechanism for embedding a window in a browser. The i4i patent seemed concrete in its initial formulation, but the interpretation of it in the i4i v Microsoft case (covering any way of separating structure and content) certainly seems as abstract as patents can get. For the NTP v RIM case, the idea of forwarding email via wireless certainly has some strong elements of abstractness. On the other hand, RIM's precise invention did bear a certain similarity to NTP's idea.

After the Mayo v Prometheus decision, the Supreme Court also sent the Ultramercial decision back to the Federal Circuit. In summer 2013 a three-judge panel of the the Federal Circuit (Rader, Lourie, O'Malley) affirmed their previous decision, though with a new argument. Judge Rader apparently felt that the abstract idea of inserting commercials into TV became non-abstract because the Internet was involved. Judge Rader referred to the following diagram outlining the interactions:

ultramercial methods

In theory, if the patent is really that specific, then workarounds should be available. In my limited reading of the patent, however, the crucial claims do indeed seem extremely broad.

Judge Lourie, while concurring, argued that the computer issue was incidental, and that overall the particular methods of the patent were concrete enough that it was not invalid due to abstractness.

That said, a month earlier the entire Federal Circuit, en banc, had created in CLS Banc International v Alice Corp a larger class of inventions that were non-patentable by reason of abstractness. In that decision, Judges Rader and Lourie were in opposition.

CLS Banc International v Alice Corp

In a decision handed down May 10, 2013, the Federal Circuit, en banc (but with only ten judges) ruled that certain computer-related business-method patents of Alice Corporation were in fact ineligible for patent protection. The patents involve the presence of a third party in a transaction, who makes guarantees so as to reduce the settlement risk for the primary parties. Having a third party mediate disputes between adversaries, and having that third party act as an escrow agent, is an ancient tradition. But the big difference, according to Alice, is that their patent is for doing this on a computer.

CLS bank was created essentially for the specific purpose of being exactly this type of escrow agent; their name is an acronym for Continuous Linked Settlement".

Alice's main patent is 7,725,375. Claim 26 is a major one. Charles Duan of Public Knowledge has claimed that the claim can be implemented via the following seven lines of code. Accounts 1 and 2 are held by Institution 1; account 3 is held by Institution 2.
    account1 = 200;
account3 = 300;
if (account1 < exchange) {print("insufficient funds"); exit(1);}
account1 -= exchange;
account3 += exchange;
print("Institution 1 should debit account 2 by " + exchange_value);
No less than seven different opinions were written. The main opinion was per curiam, or authorless; it stated that the court was split five to five and therefore the District Court's decision that Alice's patents were invalid would stand.

A major written opinion, joined by five judges total, was by Judge Lourie. Technically, it does not establish a precedent.  Lourie writes that abstract principles are not patentable; to be patentable, an abstract idea has to be applied to a concrete case.

What matters is whether a claim threatens to subsume the full scope of a fundamental concept, and when those concerns arise, we must look for meaningful limitations that prevent the claim as a whole from covering the concept's every practical application.

Lourie also took a dim view of the idea that taking a common operation and computerizing it was patentable:

[A computer is] just a calculator capable of performing mental steps faster than a human could.... Unless the claims require a computer to perform operations that are not merely accelerated calculations, a computer does not itself confer patent eligibility.

Judge Lourie tried to tie this theory back to Diehr, stating that, in that case, the use of a computer did indeed offer more than just "accelerated calculations".

It does appear, though, that Lourie implicitly expanded the scope of the "abstract ideas" exception. Recall that the Supreme Court settled the Bilski and Prometheus Labs cases based on the abstract-ideas and natural-law exceptions, respectively, but did not give a great deal of guidance in either case as to just how far these exceptions could be applied.

Judges Rader and Moore wrote in dissent:

[T]he current interpretation of § 101, and in particular the abstract idea exception, is causing a freefall in the patent system. Holding that all [claims at issue] are directed to no more than an abstract idea gives staggering breadth to what is meant to be a narrow judicial exception. And, let's be clear: if all of these claims, including the system claims, are not patent-eligible, this case is the death of hundreds of thousands of patents, including all business method, financial system, and software patents as well as many computer implemented and telecommunications patents.... This would render ineligible nearly 20 percent of all the patents that actually issued in 2011.

On June 19, 2014 the Supreme Court released their decision in Alice v CLS, and they state firmly that adding a computer to an otherwise unpatentable abstract idea can not make it patentable. Furthermore, the decision was unanimous!

Stating an abstract idea "while adding the words 'apply it'" is not enough for patent eligibility. [Mayo v Prometheus]. Nor is limiting the use of an abstract idea "to a particular technological environment." [Bilski]. Stating an abstract idea while adding the words "apply it with a computer" simply combines those two steps, with the same deficient result.

This is a clear statement that software is not eligible for patent just because it describes doing something on the computer that was formerly done another way. While this might seem obvious, note that the Federal Circuit, en banc, couldn't get a majority to agree to this.

On the other hand, the decision was not terribly deep, either. The Supreme Court still insists that algorithms are not patentable, but applications of algorithms may be. There were those who hoped that the Supreme Court would use Alice to overturn all or most software patents (though that was never realistic). Similarly, the Alice decision reinforced the position that abstract ideas are not patentable, but applications may be, all without really specifying what an "abstract idea" was.

So we still don't really know what the Supreme Court considers "abstract". The defendant did argue that an abstract idea must be a "preexisting, fundamental truth that 'exists in principle apart from any human action'", but the Court firmly and explicitly disagreed.

The decision does suggest that, in some cases, an idea may be "abstract" by virtue of being well-known, that is, obvious. This gives a new way to attack patents for obviousness, without going through the difficult rigamarole of challenging the USPTO ruling that a patent was "nonobvious". Strictly speaking, this confuses the issue of whether an invention is eligible for a patent with whether it is sufficiently non-obvious to be worthy of a patent; the two issues are addressed separately in the patent law.

Here are a few other interpretations of "abstract"

A more serious problem with the Supreme Court's doctrine that "algorithms are not patentable, but applications may be" is that it runs afoul of reality. In some deep sense, all software is an algorithm. In a practical sense, there are a lot of innovative ideas that are just a straightforward application of a particular algorithm. Compression is usually of this form; one comes up with an algorithm to describe the compression, at which point the application of the algorithm is absolutely straightforward. This would include mp3! And yet in the Bilski decision, Justice Kennedy states

the machine-or-transformation test would create uncertainty as to the patentability of software, ... and inventions based on linear programming, data compression, and the manipulation of digital signals.

Actually, all three kinds of inventions listed are fundamentally algorithms. So, according to Kennedy, the Supreme Court is not necessarily against the patenting of algorithms.

David Kappos, patent commissioner at the time of Bilski v Kappos, wrote that the Supreme Court had fundamentally upheld software patents in its Alice decision. This is not quite clear. However, reading Kappos' essay at makes it a little more clear why the USPTO continues to award such weak patents. Bilski is clearly very enthusiastic about software patents. He wrote

patentability is awarded based on the merits of the invention at hand, whether the medium in which it is built is a concrete structure, ceramic shape, metallic apparatus, molecular composition or computer language

He also wrote

U.S. software innovation is among our most important economic drivers

Alas, it now often seems to be the case that software innovation is actively threatened by the patent system.

After the Alice decision in June 2014, the Federal Circuit again revisited Ultramercial v Hulu, and on November 14, 2014 ruled that the Ultramercial "patent does not claim patent eligible subject matter" and ruled the patent invalid.

Trolls Again

The Alice decision has created problems for patent trolls; a large number of weak patents have been thrown out based on this ruling.

The Supreme Court addressed in the 2014 season two cases on when a losing plaintiff in a patent case could be required to pay court costs to the plaintiff: Octane Fitness v Icon Fitness and Highmark v Allcare Health Management System (in the latter case, Allcare held a software patent and Highmark was suing for a declaratory judgment to have it invalidated). The relevant law states this can only occur in "exceptional" cases; the Federal Circuit had earlier interpreted "exceptional" to mean if there was "material inappropriate conduct related to the matter in litigation", and not just a frivolous suit. The Supreme Court did not think much of the Federal Circuit's standard:

The framework established by the Federal Circuit in Brooks Furniture is unduly rigid, and it impermissibly encumbers the statutory grant of discretion to district courts.

And also:

The district court “is better positioned” to decide whether a case is exceptional..., because it lives with the case over a prolonged period of time.

The specter of having to pay the other side's legal fees potentially puts a huge damper on "patent-troll" litigation: it means the troll takes a much greater risk in filing a frivolous lawsuit.


Some Patent-Troll Examples

This is a touchy subject; I do not believe all patent-holding companies are "trolls". RSA, for example. Something all the trolls below have in common is that the patent claim is very weak.

Lodsys has claimed a patent on in-app purchasing (for iOS and Android apps).

MPHJ, the "scanner troll", claims to have a patent on scanning documents and emailing them to the user.

VirnetX has won a judgment against Apple; one judgment was for $625 million though the judge later ordered a new trial, which is pending. But let's look at some of the patents:
Also consider Eolas and E-Data (but possibly not I4I or NPT).

Lawsuit Venue

In recent years, plaintiffs were free to file patent-infringement lawsuits anywhere. They often chose the Eastern District of Texas.

But in 2017 the Supreme Court ruled in Heartland v Kraft that patent defendents must be sued in the state where they are incorporated. Heartland was incorporated in Indiana; Kraft had sued them in Delaware.

Many companies are incorporated in Delaware, because of its business-friendly legal environment, but patent cases are not affected by state laws.

The most popular venue for filing patent lawsuits is the Eastern District of Texas, presumably due to its plaintiff-friendly judges and juries. This is now done with forever.